Physical AI Brief
Daily cross-source signals for the Physical AI supply chain — silicon photonics, CPO, VLA models, humanoid hardware, embodied AI. Three streams, one page, zero filler.
799 items today · 729 arxiv · 8 SEC 8-K · 62 humanoid · 0 CN photonics
01 ARXIV · PHYSICAL AI PAPERS
729 items- arxiv:2611.03647 · cs.ROPixel-UMI: Learning Environment-Aware Whole-Body Manipulation from Sensorized Human DemonstrationsJason Jingzhou Liu, Tony Tao, Deepak Pathak, Ruslan Salakhutdinov
Sensorized human demonstrations offer scalable data for mobile manipulation, but embodiment differences complicate policy learning. We present Pixel-UMI, a pipeline for learning environment-aware whole-body policies from human data. We first introduce Environment-Aware Retargeting (EAR), which combines reconstructed scene geometry with demonstrated head and gripper trajectories to determine robot posture and generate collision-aware joint and base targets. Delay-calibrated action assignment converts these trajectories into actions. However, kinematic differences can leave residual pose errors,
manipulationgripper - arxiv:2611.03648 · cs.ROPixel: A Co-Designed Full-Stack Robot System for Learning General-Purpose Mobile ManipulationTony Tao, Jason Jingzhou Liu, Ruslan Salakhutdinov, Deepak Pathak
General-purpose mobile manipulation requires a combination of capable robot hardware, scalable demonstration collection, expressive whole-body control, and effective policy learning. Yet these components are often developed independently, creating mismatches that limit the capabilities of the complete system. We present Pixel, a co-designed full-stack system for learning general-purpose mobile manipulation. Pixel combines a low-cost wheeled humanoid with human-aligned kinematics and compliant grippers, multiple teleoperation and robot-free human data-collection interfaces, a unified whole-body
manipulationhumanoidteleoperationwhole-body controlgripper - arxiv:2611.03646 · cs.CVEgoActionBench: Benchmarking Dense Action Captioning for Egocentric Human VideosYujiao Shen, Jiahe Song, Haitian Li, Jingkang Yang +1
Learning from human data offers a scalable path toward embodied intelligence, yet it requires models to accurately understand fine-grained human actions from egocentric observations. However, existing benchmarks remain limited in action atomicity, temporal precision, and hand-action association. We introduce EgoActionBench, a benchmark that represents continuous human manipulation as temporally localized atomic actions associated with the left hand, right hand, or both hands. EgoActionBench contains 1,407 episodes with 41,132 human-refined action annotations. To evaluate such dense predictions
embodiedmanipulationbenchmarkevaluation protocol - arxiv:2611.03641 · cs.AIFrom Natural Language to Exact Cover: A Neuro-Symbolic Approach to Zebra PuzzlesPaulius Skaisgiris, Thomas Pammer, Veronika Semmelrock, Mykyta Ielanskyi +2
Chain-of-Thought (CoT) generation has substantially improved the performance of Large Language Models (LLMs) on complex reasoning tasks, including code generation, data analysis, and exam-style question answering. Despite these advances, purely neural LLMs continue to struggle with elementary logical reasoning problems and lack the determinism, soundness, and reliability characteristic of symbolic reasoning systems. Conversely, classical symbolic methods such as SAT solving and exact cover guarantee correctness and completeness, but require problems to be expressed in highly specialized formal
post-training - arxiv:2611.03634 · cs.LGWebHarbor: Docking Real Websites for Evolving GUI Agent EnvironmentsZhaoyang Wang, Qianhui Wu, Boyu Gou, Demi Wang +53
GUI web agents require realistic and interactive environments that remain stable during both evaluation and training. Live website evaluation suffers from changing content, access restrictions, and limited coverage of authenticated workflows, while existing synthetic environments for training often lack functional depth and visual fidelity. We propose WebHarbor, a suite of 153 executable and dockerized website environments that faithfully reconstruct the visual interfaces, content, and functional workflows of real-world web services under deterministic control. We build WebHarbor through task-
agentbenchmark - arxiv:2611.03626 · cs.CVIsekai3D SDG: A Platform for Diverse Synthetic Gameplay Visual DataRafal Karp, Gene Liu, Shih-Hsin Li, Eric Courtemanche +9
We present Isekai3D Synthetic Data Generation (SDG), a platform for simulating and capturing synthetic 3D gameplay for research in vision and world models, neural graphics, gameplay agents, and other applications. The platform combines reusable environment and character assets with Unreal Engine plugins in a two-stage workflow. First, world simulation generates gameplay while recording the world state, character movements, actions, and camera tracks needed for replay. Second, the recorded simulation is replayed to capture selected outputs, such as rendered video, images, and depth maps, at a s
world model - arxiv:2611.03625 · cs.LGFADE: Factored Ability--Difficulty Filtering for Online RLVR SchedulingTomas Duncun, Jonah Ellison, Mateo Ruiz, Hanae Fujita +1
Group-based reinforcement learning with verifiable rewards (RLVR) learns only from prompts whose sampled completions disagree: all-pass and all-fail groups produce zero relative advantage and burn generation budget without moving the policy. Adaptive schedulers try to steer rollouts toward informative prompts, yet they rely on per-prompt pass-rate trackers that entangle two quantities evolving at very different speeds---the policy's shifting competence across reasoning skills and each prompt's intrinsic difficulty. When a single exponential average or Beta counter tracks both at once, short me
memorybenchmark - arxiv:2611.03622 · cs.ROD&A: Decoupling Description and Assessment Enables VLMs as Zero-Shot Robotic Reward ModelsMarius Memmel, Shawn Yin, Urvi Rutia, Angad Wadhwa +8
Vision-Language Models (VLMs) hold promise for zero-shot video task-progress estimation in robot learning, yet existing frameworks frequently suffer from confirmation bias, hallucinating progress in failed executions due to implicit success assumptions or relative temporal ranking designs. Naive prompting of frontier models collapses into a binary success detector that fails to capture nuanced, non-linear progress and failures. While post-training approaches can mitigate this through explicit training, they demand expensive real-world robotics data and require computationally intensive post-tr
post-trainingbenchmark - arxiv:2611.03618 · cs.AIStrategist: Evolving Reasoning with AgentsTaiming Lu, Qi Xu, Zhigang Hua, Zhuang Liu
A model is only half of an AI agent system. The other half is the inference strategy: the reasoning procedure that turns model calls into a response. Today that second half is written by hand, and the experience of building one strategy never carries to the next. We introduce Strategist, a meta-agent that evolves inference strategies across tasks. Each strategy is written as a blueprint,executable code that other strategies can reuse, and together the blueprints form a library. Given a task and a small development set, Strategist retrieves related blueprints, writes candidate strategies, tests
agentai agentagent systembenchmark - arxiv:2611.03614 · cs.LGBenchmark Prediction with Sparse and Unpaired Cached ResponsesHayden Helm, Avijit Ghosh, Ben Johnson, Carey Priebe
Evaluating a generative model on all queries from all available benchmark tasks is prohibitively expensive. As such, models are often evaluated on different subsets of tasks and/or different subsets of queries from a given task. The resulting collection of responses and the corresponding (model, task) score matrix are sparse and unpaired. In this paper, we introduce the Product Kernel Perspective Space (PKPS), a collection of low-dimensional representations of models based on their embedded responses. PKPS generalizes existing methods for representing black-box generative models by comparing r
benchmark - arxiv:2611.03602 · cs.CVReviso: Not All Visual Evidence Deserves Equal Attention in Video CaptioningTengfei Liu, Yang Shi, Yuran Wang, Yuqi Tang +11
Multimodal large language models (MLLMs) can generate detailed video captions, yet they may overemphasize minor details while overlooking essential context and key events. We argue that not all visual evidence deserves equal attention, as the importance of visual information depends on the content of each video. Motivated by this observation, we propose Reviso, an agentic video captioning framework that learns what to observe and how to gather relevant visual evidence. From a low-frame-rate global preview, Reviso constructs a hierarchical Context--Focus--Points (CFP) observation plan that iden
agenticbenchmark - arxiv:2611.03601 · cs.AIDiabAgent: Toward Agentic Physiological ControlChengxi Li, Shaoxuan Li, Deyao Hong, Qingle Liu +12
As pretraining scales, medical language models have advanced rapidly in clinical question answering. Their ability to achieve effective physiological control through sustained intervention still lacks systematic quantitative evaluation. Type 1 diabetes provides a concrete setting: impaired insulin secretion creates dependence on exogenous insulin, with dosing continually adjusted using information about meals, exercise, illness, and glucose measurements. We introduce DiabAgent, a research suite for agentic physiological control comprising the Agentic Physiological Control (APC) framework, the
agentagenticbenchmark - arxiv:2611.03600 · cs.AIFour Bridges: Measuring Deception Propensity in a Multi-Agent GameRaphael Casey, Michael Simeone, Alberto Tosato, Florian Ernst +3
AI agents are increasingly deployed in settings where they interact and coordinate with other agents, yet honesty is mostly measured in single-turn, single-agent benchmarks that pressure the model to deceive. We study deception that emerges in multi-agent interaction without any instruction to deceive. We introduce Kradle, a platform for evaluating AI models, and Four Bridges, an information asymmetry game in which four agents distribute across four rooms: one is the death room, the others hold food worth more if not shared. One agent, the informed agent, knows which room is the death room and
agentai agentmulti-agentagent benchmarkbenchmark - arxiv:2611.03596 · cs.LGSpacetime Emergence in the Memory of a Learning AgentPhilip A. LeMaitre, T. Rick Perche
We show that the causal structure of spacetime can emerge within the memory of a learning agent performing a quantum information task. We train a model-based reinforcement-learning agent to maximize the quantum mutual information between localized quantum probes interacting with a quantum field in Minkowski spacetime, without directly providing it with information regarding the spacetime metric or causal structure. In learning to accomplish this task, the reinforcement-learning agent constructs an internal predictive model whose planner develops the characteristic null structure of spacetime,
memoryagent - arxiv:2611.03595 · cs.LGDynaSolver: A Neural Operator for Autoregressive 4D Flow Prediction with Dynamic BoundariesRui Qian, Shibo Dai, Mengjun Lin, Yuk-Kit Hou +5
Neural operators provide efficient surrogates for computational fluid dynamics (CFD), but most are designed for static geometries and are not equipped to model unsteady flows induced by continuously moving boundaries. We study dynamic-boundary 4D CFD prediction for articulated moving geometries and introduce RoboCFD-4D, a temporally resolved robotic CFD benchmark spanning diverse morphologies and flow conditions. To address the computational and autoregressive challenges of this setting, we propose DynaSolver, a neural operator that explicitly conditions on boundary evolution and causal flow h
benchmark - arxiv:2611.03592 · cs.LGEvaluating Uncertainty Calibration in Probabilistic Time Series Foundation ModelsThomas Decker, Volker Tresp
Time series foundation models (TSFMs) are increasingly used for probabilistic forecasting across diverse domains, yet the reliability of their uncertainty estimates remains poorly understood. In this work, we study uncertainty calibration in probabilistic TSFMs from a broader methodological and empirical perspective. We introduce a taxonomy of how TSFMs represent predictive uncertainty and develop a fine-grained evaluation framework for calibration in multi-horizon forecasting that distinguishes pooled, horizon-wise, and stronger conditional notions of calibration. Using this framework, we con
evaluation framework - arxiv:2611.03588 · cs.RO3D-BBDA: Backpropagation-Free Black-Box Domain Adaptation for Robust 3D Point Cloud RecognitionShivangi Rai, Kunal Jangid, Vaibhav Kumar, Vinod K Kurmi
Robust 3D point-cloud recognition is critical for robotics, mapping, and urban sensing, yet point-cloud classifiers often suffer performance drops under domain shifts caused by sensor changes, environmental variation, and acquisition artifacts. Many existing adaptation methods are difficult to deploy in practice because they require access to source data or the internal parameters of the source model, which may be restricted by privacy, licensing, or security constraints. We present 3D-BBDA, a backpropagation-free black-box domain adaptation framework that adapts a source-trained predictor to
memorybenchmark - arxiv:2611.03586 · cs.AIFedReady: Bounded Agents for Federation FormationZiyue Xu, Holger R. Roth, Chester Chen, Peter Cnudde
Federated learning (FL) runtimes standardize aggregation and deployment, but each new study still requires federation formation: selecting task-eligible sites and preparing their local data and task code for federated training or analysis. This preparation recurs when the target changes, while raw records and site-private mappings are excluded from cross-site FL messages. We present FedReady, a system that performs federation formation with bounded LLM agents. Agents propose shared contracts, local adapters, and task code; the runtime owns workflow state and accepts artifacts only after checks
agentllm agent - arxiv:2611.03584 · cs.ROTAMPAgent: Agentic Belief-Space Planning for Mobile ManipulationUtkarsh A. Mishra, Alperen Degirmenci, Yongxin Chen, Danfei Xu +2
Open-world robotic mobile manipulation requires acting in physical environments with only partial information. While recent Large Vision-Language Model (LVLM) agents show strong open-world reasoning capabilities, existing architectures that rely on direct tool calling struggle with geometric uncertainty and thus are limited to fully-observable tabletop environments. In contrast, Task and Motion Planning (TAMP) systems can solve long-horizon tasks under partial observability but incur high computational costs and require closed worlds. We propose \textbf{TAMPAgent}, an agentic planning framewor
manipulationagentagentictool calling - arxiv:2611.03585 · cs.ROSeeing Through the Illusion: Test-Time Scaling for World Action ModelsCheikh Ahmed, Mahdi Mostajabdaveh, Milad Yazdani, Zirui Zhou
World action models (WAMs) jointly generate actions and imagined futures, but a plausible prediction need not be realized by its co-generated action. We introduce WISE (\emph{World Action Inference Scaling through Evaluation}), a pre-execution selector that scores each sampled rollout on action validity and temporal regularity, inverse-dynamics consistency between the proposed action and the imagined motion, and instruction-conditioned task progress, and executes the highest-scoring candidate while keeping the WAM frozen. With four candidates per decision, WISE raises closed-loop success from
libero - arxiv:2611.03577 · cs.LGSonar-MoE: In-Backward Expert Updates for Memory-Efficient Mixture-of-Experts TrainingAvni Mittal, Avinash Anand, Ashutosh Kumar, Dikshant Kukreja +6
Most parameters of a Mixture-of-Experts (MoE) model belong to its experts, and standard training keeps every expert gradient in GPU memory until the optimizer step. This memory grows with the number and size of the experts, so it rises as MoE models scale. We introduce Sonar-MoE, which updates each expert tensor during the backward pass and frees its gradient right away. It brings optimizer-in-backward from dense models to batched MoE expert tensors, so the GPU never holds more than one layer of expert gradients. When we fine-tune OLMoE-1B-7B and Granite-3.1-3B-A800M in bf16 with two expert-ke
memory - arxiv:2611.03576 · cs.CLNitiBench+: A Temporal Thai Legal Question Answering Benchmark Across 17 DomainsPawitsapak Akarajaradwong, Sarana Nutanong, Chompakorn Chaksangchaichot
Law changes over time. A court decides a case under the law in force on that day, and legal QA benchmarks grade answers against today's law. We checked 156,650 decisions of four Thai courts. Up to 67% of them cite law that has since been amended or repealed. We introduce NitiBench+, the sequel of our NitiBench, a Thai legal QA benchmark with 1,670 items across 17 domains, built from decisions of three courts. For each cited section it holds both the text in force on the decision date and the current text, and its retrieval corpus holds every dated version of every Thai statute section. Answers
memoryagentbenchmarkleaderboard - arxiv:2611.03575 · cs.CLVeriHarm-Bench: Measuring Harmful Success in LLM Coding Agents VerifiablyWei-Lin Chen, Zhepei Wei, Xinyu Zhu, Yu-Min Tseng +2
Large language model (LLM) coding agents increasingly operate with shell access, allowing them to write scripts, move files, alter permissions, and modify system configurations. However, an agent may achieve a user's goal while causing unintended harm, such as deleting unrelated files, weakening access controls, or corrupting data, in response to benign inputs without explicit jailbreaking. This failure mode is particularly important because apparent task success may reassure users that an agent behaved correctly, causing harmful side effects to go unnoticed. We call this harmful success: the
agentbenchmark - arxiv:2611.03557 · cs.CVPlantVLM-Edge: Quantized Vision-Language Models for On-Device Crop Health DiagnosisGaurab Wagle, Bishal Adhikari, Jiajia Li, Xiaolei Huang +1
Timely and accurate crop health diagnosis is essential for sustainable agricultural production, yet accurate diagnosis often requires expert knowledge that is not readily available. Recent vision-language models (VLMs) have demonstrated remarkable capabilities in visual understanding, reasoning, and natural-language interaction, creating new opportunities for intelligent crop health assessment. However, deploying VLMs on resource-constrained devices remains challenging due to their substantial computational and memory requirements. This paper presents PlantVLM-Edge, a quantized vision-language
memorypost-training - arxiv:2611.03555 · cs.AIMCP Tool Poisoning in Agentic AI Frameworks: A Cross-Framework Measurement StudyFarooq Shaikh, Biagio Montaruli
The Model Context Protocol (MCP) lets agentic AI frameworks connect language models to external tool providers, but it carries attacker-controlled content (tool descriptions, schema fields, and response data) into model context without validation. Prior benchmarks measure model compliance with poisoned instructions without confirming real execution. We present the first cross-framework measurement of MCP tool poisoning with real tool execution. A 25-class root-cause taxonomy unifies prior classifications, and the open-source tool mcpvenom drives a 9,720-cell susceptibility matrix across 12 fra
agenticbenchmark - arxiv:2611.03551 · cs.AICan LLM Agents Be Forward-Deployed Engineers? Requirement Discovery and Delivery for Enterprise Text-to-SQLMarianne Menglin Liu, Rongguang Wang, Zheng Wang, Yewen Wang +4
LLM coding agents can increasingly automate model post-training, but whether they can act as forward-deployed engineer (FDE) agents, discovering customer requirements and turning them into deployment decisions, remains unclear. We study this on a real enterprise text-to-SQL engagement: a simulated customer, grounded in the real use case's requirements, starts with an ambiguous request. An agent must discover acceptance criteria, data permissions, and hardware constraints in the discovery phase, then plan and execute a post-training recipe in the delivery phase. Across 15 models, the strongest
agentllm agentpost-trainingbenchmark - arxiv:2611.03548 · cs.ROMediating Human-Swarm Interactions Through a Robot Helper: A Shepherding-Inspired ApproachArihant Yadav, Sophie Jörg, Eakta Jain
This work investigates human-swarm interaction (HSI) for steering large collectives through an intermediary robot helper. While prior robot-swarm control and HSI techniques commonly assume swarm members are addressable, these assumptions often fail for ``ungovernable'' swarms such as wildlife flocks and schools of fish. Motivated by pastoral shepherding, we compare low-level teleoperation to a higher-autonomy motion-primitives interface in a virtual reality (VR) shepherding task under varied swarm cohesion dynamics. Results indicate that motion primitives that operate on collective shape (e.g.
teleoperation - arxiv:2611.03543 · cs.LGCan LLM Agents Improve Imperfect Earth System Model Parameterizations?Kara D. Lamb
Earth system models (ESMs) rely on parameterizations to represent unresolved processes such as convection, cloud microphysics, and turbulence. However, translating advances in physical understanding into stable, validated parameterizations and integrating them into ESMs often takes years. Agentic AI could accelerate this process and enable more systematic exploration of design choices. Using warm-rain collision-coalescence as a test case, we evaluate agents' ability to synthesize scientific literature, reimplement published bulk parameterizations, and propose new functional forms as differenti
llm agentagentic - arxiv:2611.03542 · cs.LGVariational Functional Neural Networks (vfNN)Chinmay Katke, C. Nadir Kaplan
We present a variational functional neural network (vfNN) algorithm that optimizes physically motivated continuum functionals in integral form (e.g., Lagrangian, Hamiltonian, or free energy) to find the states of a system. That way, vfNN bypasses solving the equations of motion (a.k.a. the Euler-Lagrange equations), which are in the form of differential equations. To this end, our approach is fundamentally different from conventional physics-informed neural networks (PINN) that solve differential equations by minimizing a phenomenological loss function involving higher order derivatives and ad
benchmark - arxiv:2611.03534 · cs.AIAn Agentic Framework for Automated Power System Studies and Grid Connection AssessmentMaoXun Huang, Kshitij Kapoor, Eli Louise
Connecting new generation and demand to the grid depends on two scarce skills: checking applications against a large, changing body of regulation, and running the network studies that decide where and how a project connects. We present Grid Agents, an AI system that brings both together, and the methodology we use to evaluate it. The Grid Ontology is its knowledge system. It represents the regulation, standards and engineering knowledge that grid work depends on, makes it searchable, and evolves as the agents learn from use. Every answer drawn from it rests on evidence the agents cite. Grid St
ragagentagenticagent systembenchmarkleaderboard - arxiv:2611.03525 · cs.AIGrading the Unannotated: Insights from the WorkRB Skill Extraction Challenge at RecSys-HR 2026Warre Veys, Luis Gasco, Jens-Joris Decorte
Skill extraction benchmarks score models against a few human-annotated positives and treat every other skill in the vocabulary as equally wrong. At the scale of ESCO (13,939 skills) this binary view can reward models that rank the annotated positive high for the wrong reason, and hides semantically nonsensical predictions. The WorkRB Challenge at RecSys-HR 2026 replaces it with graded relevance: every candidate skill in a model's ranking window carries a 0 to 4 grade, and the leaderboard scores nDCG@100 against these grades. Since ternary judgements proved too coarse for objective annotation,
benchmarkleaderboard - arxiv:2611.03523 · cs.LGDistDebug-Bench: Can LLM Agents Diagnose Bugs in Distributed Systems?Yuxuan Jiang, Chenglin Liu, John Rose, Mason Miller +5
LLM agents show strong potential in software engineering tasks such as competitive programming and code synthesis. Existing benchmarks, however, largely focus on small-scale settings such as self-contained Python modules, leaving distributed systems, an essential class of production software, under-examined. We introduce DistDebug-Bench, a benchmark for evaluating agents' ability to diagnose complex-system failures at the code level. It curates 120 high-quality real-world bugs from 11 widely used distributed systems, spanning 5 programming languages and 10 failure patterns. Through an extensiv
agentllm agenttool-usebenchmarkevaluatorleaderboard - arxiv:2611.03520 · cs.ROPanoTac: A Dense Three-Axis Tactile Sensor for Robust Touch-Centric ManipulationBinglin, Wang, Xiaole, Jiang +6
Contact-rich manipulation benefits from tactile feedback that captures both detailed contact geometry and interaction forces. When scooping grains, contact patterns help infer spoon orientation, while resistance transmitted through the grasp guides motion. To provide these complementary sensing capabilities, we present PanoTac, a low-cost, compliant tactile fingertip designed for dense three-axis tactile sensing. Each fingertip provides a $12 \times 32 \times 3$ tactile representation, with one normal ($z$) and two shear ($x$, $y$) response components at each taxel. The sensor operates at 60 H
manipulationtactilegrasp - arxiv:2611.03517 · cs.LGAsk, Don't Equip: Question-Driven Coding Agents for Deep Kernel OptimizationGabriele Oliaro, Jaeseong Lee, Zhaoyuan Su, Sreeram Vennam +2
LLM agents for GPU kernel optimization sit at two ends of a structure-freedom tradeoff. Scripted agents fix what the model does at each step, via a learned policy, an optimization menu, or fixed roles, and are sample-efficient where the recipe applies and stall where it does not. Open-ended agents let a self-directed model mutate a population of kernels and pay for the freedom in trials: thousands of evaluations, repeated edits, and wrong candidates. Both respond to hard targets by equipping the agent with more: parsed profiler output, disassembly analysis, retrieval over documentation and lib
agentllm agentautonomous agentevaluator - arxiv:2611.03503 · cs.AIA Cross-Domain Matrix Engine for Signal Processing and Deep Neural Net InferenceMehdi Safarpour
Matrix multiplication dominates both PHY signal processing and Deep Neural Network (DNN) inference, yet these domains place opposing demands on the matrix engine. Signal processing requires low-latency execution of relatively small matrix operations, whereas DNN inference, particularly at large batch sizes, prioritizes high throughput and energy efficiency for large matrix multiplications. While offering efficient memory access profile, a Systolic Array (SA) adds fill and drain delays per matrix. Those can be amortized for large matrices but not for small matrices. Conversely, a tensor unit co
memory - arxiv:2611.03501 · cs.ROWhen Robots Outweigh Objects: How Inertia Mismatch Shapes Robot Policy LearningYounghyo Park, Antonia Bronars, Pulkit Agrawal
Conventional robot arms can be extremely heavy compared with the everyday objects they manipulate, creating a significant inertia mismatch at contact. Even a small accidental bump from a heavy arm can knock a light object out of place. Modern learning methods have nevertheless demonstrated a wide range of successful manipulation behaviors on these platforms, suggesting that even extreme inertia mismatch can be overcome with sufficient data. However, it remains unclear whether high inertia imposes a hidden cost on learning, potentially worsening scaling with data or lowering the performance cei
manipulationrobot policy - arxiv:2611.03496 · cs.AIOnline Label-Free Gradient Matching for Continual Language Model TrainingSajjad Pakdamansavoji, Mohammad Reza Rezaei, Ilija Subasic, Pedro Garzon +2
Language models in customer-facing products must be continually updated, since the data they serve shifts over time. As retraining on the full accumulated data is impractical, each update should instead run on a small set of examples carrying the training signal of all data seen so far. The strongest current methods build this set by gradient matching, which synthesizes examples whose training gradients match those of the full data. However, they require ground-truth labels, run once on a fixed snapshot rather than on a growing stream, and are slow on large data. We introduce \textbf{OLGA}, wh
memorybenchmark - arxiv:2611.03494 · cs.AIHarder Than the Task: From Technical to Organizational Difficulty in AI BenchmarkingVictor Barres, Aden Barton, Nicolai J. Foss, Henrik M. Nielsen
AI benchmarks have made remarkable progress in measuring the ability of AI systems to perform increasingly demanding cognitive and technical tasks. Yet, many of these benchmarks treat technical tasks as the unit of analysis, while much economically valuable work takes place within organizations, where tasks are embedded in a system of relations. Within organizations, successful task performance depends not only on solving the technical demands of the task but also on navigating the interdependence of activities, distributed knowledge, differentiated authority, temporal constraints, conflicting
ai agentbenchmark - arxiv:2611.03489 · cs.ROExploring Proximity and Tactile Sensing for Robotic ManipulationEric T. Chang, Bianca Aumann, Zhanpeng He, Mohamed Elnoor +6
Pre-touch sensing, or sensing information about an object before contact, is an attractive modality for robot hands as it can allow a robot to anticipate and respond to a contact before it happens. In this work, we introduce ToFTouch, a multimodal tactile sensor combining three multizone (4$\times$4 resolution) time-of-flight sensors with an optics-based pressure sensing array. We mount this sensor on a parallel jaw gripper and demonstrate its utility on three distinct tasks: (i) pre-grasp adjustment, (ii) item extraction from clutter, and (iii) item insertion into tight spaces. For pre-grasp
manipulationtactilegrippergrasp - arxiv:2611.03465 · cs.CVHIVE-Diff: High Frame Rate HDR Video Reconstruction from a Single LDR Frame and EventsCaixin Wang, Qingrui Wang, Jie Zhang, Ralph Etienne-Cummings
High frame rate HDR video is important in applications such as autonomous driving and scientific imaging, where rapid motion and extreme illumination often occur simultaneously. Conventional frame cameras face a trade-off between temporal resolution and dynamic range, while event cameras provide complementary high-speed brightness-change measurements over a wide illumination range. These properties have motivated event-based HDR reconstruction, with existing methods broadly following event-only and hybrid event-frame paradigms. Event-only HDR reconstruction, however, suffers from brightness an
event camera - arxiv:2611.03446 · cs.ROClinician-AI Collaboration in Endovascular Neurointerventions: In-Vitro User Study of Copilot Assistance Modes for Robotic Navigation Across the Autonomy SpectrumNikola Fischer, Harry Robertshaw, S. M. Hadi Sadati, Xingyu Chen +15
Mechanical thrombectomy (MT) is an effective treatment for acute ischemic stroke, yet access remains limited by the availability of neurointerventional expertise. Telerobotic and AI-assisted systems may help address this challenge, but it remains unclear how clinicians interact with different levels of autonomy and how autonomous navigation systems perform in realistic environments. This work introduces Copilot Assistance Modes (CAMs), a framework for clinician-AI collaboration in robotic endovascular navigation spanning the autonomy spectrum from manual operation to full autonomy. In the firs
sim-to-real - arxiv:2611.03434 · cs.AIOptimize the Harness, Not the Model: Fitting a Minimal Agent to a Self-Hostable LLM for Investigative AnalysisShai Nahum Gefen, Vered Chassida, Dmitry Basin, Guy Feigenblat
Government intelligence teams need LLM agents that can analyze large, messy evidence collections, but security and cost constraints often rule out hosted frontier models and broad, hard-to-audit tool surfaces. We report CLEO: Constrained Loop for Environment Optimization, a method for fitting a minimal agent harness to a fixed, self-hostable LLM. Given an expert-validated dataset, CLEO admits only the tools, guardrails, memory policies, and loop components that improve accuracy while remaining general and auditable. On our held-out dataset, the CLEO harness reaches an 83.8% pass rate, within 1
memoryagentllm agent - arxiv:2611.03433 · cs.LGA Living Leaderboard for Graph Machine LearningMartin Ritzert
Graph machine learning benchmarking results are scattered across papers that use quite different sets of datasets and baseline methods, making it difficult to tell whether a reported gain reflects architectural progress or merely an outdated comparison or weakly tuned baseline. We introduce a living, graph-learning-specific leaderboard that organizes reported results by dataset, putting new results into context. As of 15 October 2026 our LLM-based extraction pipeline extracted 27,723 eligible results across 111 datasets from 885 peer-reviewed papers. Our analysis of these results quantifies co
benchmarkleaderboard - arxiv:2611.03430 · cs.ROEvaluating Depth-Augmented Visual Features for Cross-Dataset Lunar Traversability EstimationJan Baturo, Bartosz Ptak
Autonomous mobility is essential for sustained lunar surface operations, as communication delays and line-of-sight occlusions limit continuous teleoperation. Although recent traversability estimation methods increasingly rely on deep learning, the domain gap between terrestrial analogue environments and real lunar terrain can limit their generalisation, and real lunar traversability labels are scarce, making fully supervised training difficult. To address these challenges, we build a weakly supervised framework trained on sparse pairwise annotations and study whether augmenting visual features
teleoperation - arxiv:2611.03428 · cs.ROToward Uncertainty-Aware, Physics-Grounded Human World ModelsYufei Zhang, Zijun Cui
A robot that acts around people needs a world model of the human: a belief over the person's body state and a dynamics model that predicts where that state goes next. This abstract argues that two published components form exactly these two halves, and specifies how to couple them. KNOWN [1] reconstructs the 3D human body while estimating both aleatoric and epistemic uncertainty in real time (11.6 ms per frame); its aleatoric term identifies unreliable observations, its epistemic term identifies inputs from low-density regions of the training data, and refinement driven by the epistemic term r
world modelevaluation protocol - arxiv:2611.03424 · cs.CLIdeaTaste: Do LLMs Share the Research Taste of Human Researchers?Shanshan Zhong, Chen Xu, Jingjie Ning, Zhengzhong Liu +2
Research taste, the judgment of what to read, which problems to pursue, and which methods to try, increasingly shapes the work of AI scientists, yet it remains far less understood than their ability to execute research. We introduce IdeaTaste, a benchmark that scores an LLM by how closely the research proposals it writes from a literature snapshot frozen at 2025 match the work the same communities published in 2026. Our experiments span six contemporary language models and four research areas from established to emerging fields, covering more than 90{,}000 proposals in total. LLM agents fall w
llm agentbenchmarkleaderboard - arxiv:2611.03420 · cs.LGTemporal Island Sparse Autoencoders for Interpreting Clinical Time-Series ModelsMuhan Yeo, Bong Gyun Kang, HyunGi Kim, Sungroh Yoon
Deep learning models for electronic health records achieve strong predictive performance but provide little insight into what clinical concepts they represent internally. Post-hoc attribution methods produce per-cell importance scores that lack temporal coherence, cross-variable structure, and reusability across patients. We introduce the Temporal Island Sparse Autoencoder (TI-SAE), which learns a backbone-specific, patient-shared clinical concept dictionary from a frozen predictive model's representations. Each dictionary entry (latent) is defined by learned temporal Gaussian islands (specify
benchmark - arxiv:2611.03414 · cs.LGFinite-Memory Control of POMDPs: Fundamental Limits and Efficient DesignEmrecan Kutay, Atilla Eryilmaz, Ness Shroff
Optimal control of partially observable Markov decision processes (POMDPs), in general, requires a complete history. The belief state, a distribution over hidden states, summarizes this history and provides the classical sufficient statistic to control POMDPs. However, the continuous nature of the belief space generally prevents a finite-memory controller from maintaining it exactly. This raises the question we address in this work: what must the available finite memory preserve to achieve near-optimal control performance? We study this question through an information-theoretic converse that g
memorybenchmark - arxiv:2611.03412 · cs.AIDeliverBench: Benchmarking AI Agents on Delivering Realistic Multi-Stage Professional ProjectsYumou Liu, Mingju Chen, Zhoufutu Wen, Qianyu Yang +17
LLM agents are judged by what they deliver, and in professional work what gets delivered is a project, not a task. A project pays off only if the whole package holds, and one weak part can limit the value of the rest. Yet existing benchmarks mostly test agents on one task at a time, leaving a gap between the capability we measure and the outcome that makes agents useful. To address this gap, we introduce DeliverBench, which tests whether agents can carry a project from brief to final delivery. Its 11 projects, spanning financial due diligence, legal contract drafting, and drug safety reporting
ai agentllm agentbenchmark - arxiv:2611.03411 · cs.LGLow-Rank Clebsch--Gordan Coupling for Equivariant Spherical ForecastingYaojun Li, Zhixuan Zhao, Dongzhe Zheng, Christine Allen-Blanchette
Spherical-harmonic neural operators provide an efficient representation for global forecasting, but their discretized pointwise nonlinearities can break rotational equivariance through aliasing. Clebsch-Gordan (CG) tensor products provide exactly equivariant interactions between spherical-harmonic representations, but their computational cost grows rapidly with spectral bandwidth and has limited their use at forecasting scale. We introduce LCG-SFNO, which combines low-rank CG interactions within a controlled low-degree band with coefficient-space residual propagation and an exactly projected q
benchmark - arxiv:2611.03406 · cs.AIFrom Reactive to Agentic Resilience: Multi-Timescale AI-RAN Orchestration Under Signaling StormsSotiris Chatzimiltis, Mahdi Boloursaz Mashhadi, Mohammad Shojafar, Rahim Tafazolli
Next-generation open radio access network (Open RAN) and AI-RAN deployments turn the RAN into a disaggregated, compute-aware platform whose service continuity depends on many control loops acting at different timescales. Resilience in this setting is best seen as a lifecycle spanning anticipation, absorption, adaptation, recovery, and evolution. Yet, existing mechanisms remain stage-specific and are rarely coordinated across the control hierarchy. This paper argues that agentic AI offers an orchestration layer that coordinates this lifecycle, turning operator intent into coordinated action whi
agentagentic - arxiv:2611.03404 · cs.CVVidDub: Replacement-Aware Distillation for Two-Step Video DubbingYexin Liu, Yunlong Lin, Zehan Wang, Guangxin He +10
Video dubbing aims to synthesize lip movements that match target speech while preserving the speaker identity, head pose, and background of a source video. Existing editors are predominantly trained via self-reconstruction under original paired audio, even though the source mouth becomes a conflicting target when driving speech is replaced. We introduce replacement-aware cross-audio supervision: paired-audio samples use full-frame reconstruction, whereas unpaired replacement-audio samples mask out source-mouth reconstruction losses and supervise the mouth solely through audio-visual synchroniz
benchmark - arxiv:2611.03399 · cs.CVImage Guided Depth Upsampling using Anisotropic Total Generalized VariationDavid Ferstl, Christian Reinbacher, Rene Ranftl, Matthias Rüther +1
In this work we present a novel method for the challenging problem of depth image upsampling. Modern depth cameras such as Kinect or Time of Flight cameras deliver dense, high quality depth measurements but are limited in their lateral resolution. To overcome this limitation we formulate a convex optimization problem using higher order regularization for depth image upsampling. In this optimization an anisotropic diffusion tensor, calculated from a high resolution intensity image, is used to guide the upsampling. We derive a numerical algorithm based on a primaldual formulation that is efficie
benchmark - arxiv:2611.03393 · cs.CVTaoMate-FVSR:Audio-Guided Facial Video RestorationXinyang Li, Zeran Xu, Qijun Gan, Chenwei Zhang +2
AI-generated videos often depict people within detailed scenes while leaving their small facial regions poorly resolved. Visual super-resolution can sharpen these regions without resolving whether the recovered lips and teeth agree with the spoken audio. We present an audio-guided facial video restoration framework that addresses this ambiguity within a reference-guided video flow model. Source--noise mixing initializes reconstruction from a partially preserved observation, and structured reference injection separates appearance anchors from continuous motion context. Training combines restora
benchmark - arxiv:2611.03390 · cs.LGProgressive Multi-Scale Super-Resolution via Cascaded Bicubic Anchors and Inter-Scale Mutual RefinementChun-An Lin, Yu-Chia Chang, Chi-Mao Fan, Yen-Chieh Ouyang +1
Super-resolution networks are trained for one magnification factor, so several output resolutions require several models. We propose PMS-SR, a progressive single-input multiple-output network that produces x2, x4, and x8 results in one forward pass, with a Transformer backbone on the low-resolution grid. In Stage 1, each scale is predicted as a residual over the detached bicubic upsampling of the previous scale's prediction. With Dirac and zero initialization, training starts exactly from the bicubic cascade, and each head adds only the band the previous scale lacks. In Stage 2, inter-scale mu
benchmark - arxiv:2611.03384 · cs.AICan We Stop the Backtest Yet? Anytime-Valid Promotion Tests for Adaptive Financial Research AgentsEric Benhamou, Jean-Jacques Ohana, Beatrice Guez, Ethan Setrouk +2
Financial research agents can generate and monitor trading rules faster than a human can document the resulting selection process. Candidates appear at data-dependent times, weak rules are abandoned, and promising ones are watched until they cross a reporting threshold. We introduce PROMOTE, a protocol that assigns every candidate predictable testing wealth at birth and accumulates a cost-aware e-process only from subsequent returns. If lifetime allocations sum to at most one, a weighted Ville argument bounds the probability of ever promoting a rule whose declared bounded net score has always
agent - arxiv:2611.03382 · cs.CVASCoR: Asymmetric Semantic-Spatial Learning for Remote Sensing Domain GeneralizationZitong Wang, Shuchang Lyu, Meng Li, Qiang Wang +2
Source-only domain generalization (DG) for remote sensing semantic segmentation seeks reliable predictions in unseen domains without target-domain data during training. Although vision foundation models offer transferable knowledge, category recognition remains sensitive to source-specific visual cues, while task-specific adaptation may weaken pretrained transferability. We propose ASCoR, an asymmetric semantic-spatial framework built on SAM3. Its text-conditioned semantic branch uses class-name prompts to provide explicit category guidance. Probe-Once Dynamic Adaptive LoRA distributes a fixed
benchmark - arxiv:2611.03378 · eess.SYNon-Coherent Over-the-Air Distributed Consensus: Algorithm Design and Power Control OptimizationYuhang Deng, Zheng Chen, Erik G. Larsson
Non-coherent over-the-air (NCOTA) aggregation offers a scalable and communication-efficient approach to distributed average consensus in wireless networks, without requiring instantaneous channel state information (CSI) or phase alignment. However, persistent aggregation errors pose fundamental challenges in achieving exact average consensus under nominal distributed averaging protocols. In this paper, we develop an NCOTA consensus framework that accommodates row-stochastic mixing, without requiring the conventional average-preservation property of doubly stochastic mixing matrices. Building o
benchmark - arxiv:2611.03377 · cs.CVSame Outputs, Different Scores: Auditing the Unchecked Graders of Multimodal LLM BenchmarksAhmad Jarjis Hasan, Jonathan Lwowski
Every multimodal-LLM multiple-choice benchmark ends with a scoring rule that converts a model's response, often free-form text rather than a clean option letter, into a right/wrong verdict, buried in its code. Unlike the dataset, this rule is chosen without justification and left unexamined, yet it sets the published number. We reproduce the official scorers of eight peer-reviewed benchmarks on five open vision-language models (VLMs) that fit a single 16 GB GPU, and introduce a blinded human-plus-LLM-judge adjudication (Cohen's kappa=0.956) that recovers what the model actually answered. The s
benchmarkleaderboard - arxiv:2611.03374 · cs.LGCharacterizing When Language Models Take Harmful Actions to Protect Their Assigned GoalsBrandon Jaipersaud, Elliot Creager
Recent language model (LM) safety failures have largely been studied as disparate phenomena. In agentic misalignment, a model blackmails an executive to avoid being replaced. In alignment faking, it complies with harmful requests during training to keep its values from being modified. We argue that these behaviours share a common decision structure: a model assigned a goal must either take a harmful action to preserve it or refuse and lose it. We formalize this as a goal-harm conflict and present the same conflict under various elicitation protocols that range from compact forced-choice prompt
agentictool use - arxiv:2611.03371 · cs.CVEO-Jev: Decision Model for Earth ObservationMohamed Atef Abouelhadid, Diaa Addeen Abuhani, Muhammad Akhtar Munir, Fahad Shahbaz Khan +1
Earth observation (EO) workflows require selecting among known outcomes, estimating confidence, and processing large volumes of imagery. Generative vision-language models can address these tasks, but autoregressive decoding is computationally costly and their confidence estimates are unreliable. Decision models offer an alternative, returning probabilities over candidate answers in a single forward pass. However, their accuracy, calibration, and transfer across EO tasks remain insufficiently understood. We investigate these properties through a diagnostic study of d3-flash (8.4B parameters) ac
benchmark - arxiv:2611.03370 · cs.CVFuse To Align: Multimodal Representation Learning for Retrieval via Hypergraph RegularizationAnindya Nag, Ambuj Mehrish, Sebastiano Vascon
Multimodal retrieval must integrate complementary evidence from views that may be missing, noisy, or unequally informative. When a sample provides limited reliable evidence, learning from its views alone can hinder the acquisition of useful retrieval representations. Semantically related training samples may provide complementary observations, yet existing volume-based objectives jointly align modalities within each sample without explicitly enabling information exchange across samples. We therefore propose a training-time hypergraph fusion framework that connects observed modalities both with
benchmark - arxiv:2611.03350 · cs.LGCertifying Graph Neural Networks Against Label and Structure PoisoningLukas Gosch, Xichuan Chen, Yan Scholten, Stephan Günnemann
Robust machine learning for graph-structured data has made significant progress against test-time attacks, yet certified robustness to poisoning - where adversaries manipulate the training data - remains largely underexplored. For image data, state-of-the-art poisoning certificates rely on partitioning-and-aggregation schemes. However, we show that these methods fail when applied in the graph domain due to the inherent label and structure sparsity found in common graph datasets, making effective graph-partitioning difficult. To address this challenge, we propose a novel semi-supervised learnin
benchmark - arxiv:2611.03344 · cs.LGFLAMIA: FLAsh MIcroscaling Attention for Fast and Causally Safe Training on BlackwellFedor Velikonivtsev, Mikhail Koltakov, Denis Kuznedelev
Attention remains a substantial computational cost in Transformer training, particularly at long context lengths. NVIDIA Blackwell GPUs provide high-throughput low-precision matrix multiplication, but exploiting it in attention requires addressing quantization overhead, causal information leakage through shared scales, and numerical inconsistencies between the forward and backward passes. We present FLAMIA, a family of attention kernels for Blackwell with MXFP8 and mixed NVFP4/FP8 training paths. Our design combines token-block scaling that reuses quantized operands across matrix orientations,
long context - arxiv:2611.03339 · cs.CVArabGeoText: Tool-Augmented Vision-Language Models for Arabic Geospatial Text SpottingYakoub Bazi, Sarah Alsaleh, Mohamad M. Al Rahhal, Mohamed A. Mekhtiche +1
Geospatial text spotting involves locating and transcribing geographic labels, such as road names, district names, and landmarks, in raster maps. These labels are often small, densely arranged, and arbitrarily oriented. Although Arabic text recognition has been studied extensively in documents and natural scenes, Arabic geospatial text spotting remains underexplored. We introduce ARABGEOTEXT, an Arabic-centric benchmark covering 970 geographic locations in Riyadh and Jeddah, with 17,875 manually annotated text instances per view type. For each location, we pair an OpenStreetMap (OSM) rendering
benchmark - arxiv:2611.03330 · cs.LGThe SIMPLE Law for Energy Cost of LLM InferenceHiari Pizzini Cavagna, Andrea Proia, Daniele Cesarini, Zeynep Kiziltan +1
Generative AI is billed per token, yet tokens are not equally expensive. Each token enters the KV-cache and raises the cost of those that follow, so energy grows non-linearly along a sequence whose tokens look identical. No analytical model states what a token costs, and the most accurate SOTA model is a fit, not a law: it absorbs architecture into fitted coefficients, requiring per-LLM fitting. We derive closed-form expressions of the computational and memory-access complexity of LLM inference in terms of architectural hyperparameters, covering MHA and GQA, and compare a family of analytical
agentic - arxiv:2611.03327 · cs.AIFrom FAIR to AI-FAIR Metadata-First Infrastructure for AI-Enabled ScienceElizabeth G. Campolongo, Eric R. Sokol, Marty Downs, Gina Glenne +8
Artificial intelligence (AI) technologies and their capabilities have advanced rapidly over the past few years, leading to widespread interest and adoption in the biological and ecological sciences. However, the data infrastructure and computational resources that provide the foundation for AI workflows in ecology and biodiversity science remain fragmented. Moreover, the flow of value remains largely unidirectional: data move from collectors, through repositories and aggregators into models, and finally to research products or management actions. The derived value, such as corrected taxonomies
agentic - arxiv:2611.03319 · cs.AITrafficInstruct: A Multi-Task Network Traffic Dataset and Benchmark for Instruction TuningZhengdao Li, Hui Wen, Zixuan Wan, Weifeng Zhang +5
Instruction tuning for multi-task network traffic analysis requires data with clearly defined sources, supervision targets, and evaluation conditions. We introduce TrafficInstruct, a newly collected dataset comprising raw packet captures and 55,000 structured supervised fine-tuning records across ten tasks: malware traffic detection, botnet detection, malicious DNS-over-HTTPS detection, Web attack detection, advanced persistent threat detection, encrypted VPN detection, Tor behavior detection, encrypted application/service classification, website fingerprinting, and application classification
benchmark - arxiv:2611.03309 · cs.CLMultiInpaintCOCO: Fine-Grained Visual Understanding in Multiple LanguagesAndrei-Alexandru Manea, Jindřich Libovický, Philipp J. Rösch
Previous research shows that vision-language (VL) models struggle to understand fine-grained image details and tend to focus on the overall layout. Moreover, any weakness in English tends to be further amplified in the multilingual setting. To study that, we extend InpaintCOCO, an English dataset focused on fine-grained visual details, to 13 typologically diverse languages across five language families, creating MultiInpaintCOCO - a new multilingual benchmark of 953 minimal-pair examples. We further test several state-of-the-art models in multilingual multimodal retrieval, considering both lig
benchmark - arxiv:2611.03308 · cs.LGLLMs Depend on Specialist Tools for Out-of-Distribution Generalization in MS/MS Structural ElucidationGabriel Asher, Cailum M. K. Stienstra, Tomo Oga, Devesh Shah Antonio H. de O. Fonseca +2
A frontier language model was recently reported to solve 1D NMR structural elucidation without tools. We ask whether this capability transfers to tandem mass spectrometry (MS/MS) in the de novo setting, where the model must propose a structure directly from the spectrum without a candidate list or reference library. To separate reasoning from recall, we build a 100-analyte benchmark from MassSpecGym, split by each analyte's textual footprint into in-distribution (well documented in literature) and out-of-distribution (essentially undocumented) arms. We hold the model, harness, and output forma
agentbenchmark - arxiv:2611.03301 · cs.LGAn Axiomatic Framework for Reinforcement Learning in Open Multi-Agent SystemsJianhong Wang, Yang Li, Samuel Kaski, Jonathan Lawry
Open multi-agent systems, where only a subset of agents are controllable, are central to real-world domains such as smart grids, yet lack a principled foundation for reinforcement learning algorithm design. We study this setting through n-agent ad hoc teamwork (NAHT), a recent framework for learning in open multi-agent environments. We propose an axiomatic framework that derives value function learning from the Shapley axioms (Efficiency, Symmetry and Linearity), treating them as structural constraints rather than heuristic design choices. This formulation builds on a novel NAHT decomposition,
agentmulti-agentagent systembenchmark - arxiv:2611.03298 · cs.AISQLoom: Set-Contrastive Learning of Syntactic, Semantic, and Logical Dependencies for Text-to-SQLHao Fu, Degang Sun, Wan Wei, Jinxia Wei +2
Text-to-SQL translates natural language questions into SQL queries, enabling database access without writing SQL. Fine-tuning open-weight large language models (LLMs) is cost-effective, but existing methods treat SQL as ordinary text and optimize token-level likelihood without explicitly modeling its syntactic, semantic, and logical requirements. Thus, a single incorrect alias or comparison operator can invalidate a query while contributing little to training loss. We propose SQLoom, a constraint-aware fine-tuning framework with supervision at three levels: syntactic validity under SQL grammar
agentllm agent - arxiv:2611.03294 · cs.ROForceLDM: A Force-aware Latent Dynamics Model for Contact-Rich ManipulationPeiyuan Tang, Yao Li, Xiaodong Zhang, Wuyang Zhang +3
Contact-rich manipulation requires fine-grained control and complex physical interactions with the environment, posing significant challenges for robot policy learning. Vision-only policies struggle with such tasks, as contact states and interaction forces are difficult to infer from visual observations alone. Existing methods mitigate this issue by incorporating force or tactile sensing. However, they often treat these signals as auxiliary observations without fully exploiting their role in modeling future interaction dynamics. To address this, we propose ForceLDM, a force-aware latent dynami
manipulationtactilerobot policylatent dynamics - arxiv:2611.03293 · cs.LGENGINE: Endogenous Variational MultiScale Optimization for Zeroth-Order LLM Fine-TuningZhuoli Ouyang, Changxi Chi, Siyuan Li, Tailin Wu
Zeroth-order (ZO) optimization enables memory-efficient large language model (LLM) fine-tuning, and recent methods incorporate curvature-aware strategies. However, in the high-dimensional parameter space of LLMs, these methods inherently suffer from exploding gradient variance and prohibitive costs. Conversely, confining updates to a low-dimensional subspace reduces variance but discards critical high-dimensional information, degrading performance. This creates a fundamental dilemma: full-space optimization is informative but has high variance, while subspace optimization has low variance but
memorybenchmark - arxiv:2611.03289 · cs.AITop-$k$ Retrieval Leaves Memory Value UndeterminedArman Behnam
A memory-augmented agent must decide which memories to keep, merge and retrieve, and its only evidence of their value is its retrieval log: the memories it retrieved in each episode and whether it succeeded. Existing methods assign every memory a value from this log, even when the log cannot separate one memory's contribution from another's. We treat the log as an experiment that nobody designed and ask which memory values it determines. Two counts answer this before any outcome is observed: how often each memory was retrieved, and how often each pair was retrieved together. From these counts,
memoryagentbenchmark - arxiv:2611.03283 · cs.CVReVA-HPE: Relative 6DoF Head Pose Estimation with Synthetic Pose Supervision and Personalized Virtual AnchorsVasiliki Vasileiou, Panagiotis P. Filntisis, Petros Maragos, Kostas Daniilidis
Monocular head pose estimation is usually formulated as absolute regression from a single face crop, yet many applications observe the same person over time and can supply an explicit anchor image. Recently, relative head pose estimation was proposed, which casts the problem as estimation of the relative geometric change between an anchor and a query image. However, existing methods still require the anchor image pose to be known at inference, significantly limiting their practical use. We first extend relative head pose estimation to full 6DoF by additionally predicting a depth-normalized rel
benchmark - arxiv:2611.03268 · cs.ROVerifLM: Physically Verified Image-Edit-Guided Goal Pose Generation for Robotic ManipulationXuanyou Chen, Jinyu Chen, Glen Chou
Many manipulation goals are easier to visualize than to specify numerically through explicit geometric constraints. We present VerifLM, a framework that converts image-edited goal configurations into physically verified 6-DoF object poses for planning. Using an edited image, VerifLM recovers an initial relative pose between an active object and a passive object, then prompts a vision-language model to generate axis-wise sampling constraints and a task-specific validation program. Candidate poses are evaluated in a physics simulator, and the VLM selects the most task-consistent retained candida
manipulation - arxiv:2611.03260 · cs.LGThe Cost of Mismatch: Noise Amplification in Zeroth-Order Reinforcement LearningLianmin Chen, Junbin Qiu, Chenxing Wei, Yao Shu +1
Zeroth-order (ZO) optimization has emerged as a promising approach for memory-efficient large language model (LLM) post-training, yet its reliable application to reinforcement learning with verifiable rewards (RLVR) remains largely unexplored. Motivated by the distinct data regimes of on-policy RL and fixed-data optimization, we develop a generic variance-theoretic framework to investigate what limits the reliability of ZO gradient estimation. Our analysis reveals that estimation reliability is structurally determined by whether the two finite-difference evaluations are performed on matched da
memorypost-trainingbenchmark - arxiv:2611.03250 · cs.LGAnomaly-ranked confirmation slows new-class learningArgyro Sasli, Michael Coughlin
Anomaly detectors rank unfamiliar examples for expert confirmation, and the confirmed examples teach a classifier to recognise a new class. Across five benchmarks from physics, medicine and computer vision, the members flagged by the detector are those the classifier learns first. In three systems, confirming the highest-ranked members of a new class gives lower recognition than confirming the same number of random members. Handing subsequent selection to the promoted classifier recovers part of this loss in two systems. A paired intervention further shows that replacing nearby known classes w
benchmark - arxiv:2611.03247 · cs.CVR$^2$-Omni: Reducing Structural and Precision Redundancy in Omni-LLMs for Long-Form GenerationBolong Feng, Pengfei Zhou, Zhenglin Wan, Wei Wang +1
Omni-modal large language models (Omni-LLMs) typically process large numbers of multimodal tokens from audio-visual inputs, incurring substantial computational costs and a large memory footprint for the key-value (KV) cache. Aggressive token pruning can alleviate this problem while maintaining competitive performance on multiple-choice tasks, which require only a single option as output. However, we observe substantial quality degradation when generating long-form, open-ended responses. We attribute this gap to query-guided selection, which can retain sufficient evidence for answer selection w
memorybenchmark - arxiv:2611.03244 · cs.LGForecasting How RLHF Responds to Reward-Model Edits Before TrainingZixiao Wang, Duzhen Zhang, Xingbo Du, Dongxia Wu +1
Responses with headers, lists and greater length receive high scores from reward models (RMs), and RLHF policies learn to produce these features even when true quality does not improve. Practitioners therefore edit their RMs, but they learn whether an edit works only after an RL run. We show that the response of a policy to an RM edit can be forecast before training, from samples of the current policy and the scores of the old and new RM. If we read the RM as a set of prices, the first-order theory of KL-regularized RL implies that each feature moves in proportion to its within-prompt covarian
rlhf - arxiv:2611.03238 · cs.AIScienceCUA: Resource-to-Skill Construction and Execution-Guided Skill Evolution for Scientific Computer Use AgentsDaiqiang Li, Jingyi Peng, Rui Hu, Jia Li +6
Computer-use agents for scientific software require application-specific expertise. Much of this expertise is documented in human-authored software manuals and web pages. However, turning these resources into executable, reusable skills remains challenging due to mismatches between documented and deployed software versions and differences in scope between documentation coverage and task requirements. We introduce ScienceCUA, a two-stage framework combining resource-to-skill construction with execution-guided skill evolution. First, ScienceCUA identifies tasks supported by the software from its
benchmark - arxiv:2611.03233 · cs.AIAxion 1.0: Open Agent Foundation Models, SimplifiedYuqi Jia, Yunjie Tian, Jiayu Zheng, Yijun Liang +3
We introduce Axion 1.0, a fully open agent foundation model that combines strong and broad agentic capabilities with a remarkably simple training recipe. Axion 1.0 is trained in only three stages: two lightweight supervised fine-tuning stages followed by reinforcement learning, without dedicated mid-training, direct preference optimization (DPO), or on-policy distillation (OPD). The first SFT stage develops strong self-contained reasoning through model-aware data construction, where we synthesize challenging problems and adversarially strengthen them based on the base model's own successful so
agentagenticbenchmark - arxiv:2611.03229 · cs.CVCL3DOD: Understanding the Stability-Plasticity Dilemma in Outdoor Class-Incremental 3D Object DetectionMuhammad Ahmed Ullah Khan, Khalil Ahmed, Didier Stricker, Muhammad Zeshan Afzal
Class-incremental learning (CIL) for outdoor 3D object detection remains largely underexplored. Existing outdoor studies are limited to single-benchmark evaluations, with SDCoT++ considering KITTI and CL2OD considering nuScenes under different incremental settings. This makes it difficult to assess whether observed continual-learning behavior generalizes across autonomous-driving benchmarks. To the best of our knowledge, we provide the first unified cross-benchmark analysis of outdoor class-incremental 3D object detection across multiple autonomous-driving benchmarks, covering both two-task an
benchmark - arxiv:2611.03227 · cs.LGWhat Structurally Breaks an Agent? Auditing Difficulty in Stateful Tool-Use WorkflowsShereen Anand, Ju Wang, Yali Du
An aggregate task score does not distinguish failures of scheduling, state retention, and candidate selection. We develop a small executable instrument that varies dependency depth, candidate count, active state width, and access to current-state snapshots while holding the number of required operations fixed. Two dependency-graph families show that increasing depth at fixed work changes both scheduling freedom and the number of reachable completion states, so these controls are coupled, and we quantify the coupling. The primary study contains 49,152 generated tasks and 393,216 reference-contr
tool-use - arxiv:2611.03223 · cs.AIDo Agents Still Follow the Rules? Benchmarking Rule Compliance under Agent Context CompactionYiwei Zha, Qingyu Meng, Herbert Bos, Min Chen
LLM agents compact long histories to fit their context windows, but the compacted context must still carry the operating rules that govern later actions. Prior work shows that violations rise after compaction, but measures compliance outside a complete agent harness and does not identify where each violation comes from. We introduce ARCBench(Agent Rule Compliance Benchmark), a configurable benchmark built on the open-source pi agent harness. Its tool-calling agent loop lets us measure rule compliance from actions taken during execution, bringing the evaluation closer to deployed agent workflow
agentllm agentbenchmark - arxiv:2611.03219 · cs.AICan Coding Agents Sustain Continual Evolution without Forgetting?Shangziqi Zhao, Yitong Zhang, Jia Li
Recently, harness self-evolution has emerged as an important paradigm for improving agent performance. Prior studies show that evolving the agent harness can effectively enhance agent capabilities without modifying the foundation model's parameters. We find that existing approaches largely study agent evolution on a single task. In practice, coding agents may need to handle multiple task types over time. Whether an agent harness can continuously evolve across multiple tasks without suffering from catastrophic forgetting remains an important open question. In this paper, we investigate this ope
agent - arxiv:2611.03209 · cs.LGDynSL: Resource-Adaptive Model Training across Heterogeneous Edge DevicesBoris Radovič, Istemi Ekin Akkus, Marco Canini, Ruichuan Chen +3
Training models over data distributed across edge devices presents a trade-off: centralizing data simplifies training but raises privacy and regulatory concerns, while full local training may exceed the devices' compute and memory capacity. Split learning addresses this challenge by partitioning the model between clients and the server, allowing part of the computation to be offloaded. However, existing approaches typically rely on static split configurations or simplified assumptions that fail to capture the heterogeneity and complex dynamics of real deployments. We present DynSL, a framework
memory - arxiv:2611.03200 · cs.LGCISA: Retrofitting Dense Attention Models with Chunk-Indexed Sparse AttentionMohammed Dabbah, Izhak Golan, Oren Tropp, Daniel Afrimi +9
Long-context LLM inference is costly because attention repeatedly reads a growing key-value (KV) history. We introduce Chunk-Indexed Sparse Attention (CISA), which retrofits pretrained dense attention models without gradient updates. CISA selects fixed-size KV chunks using compact summaries of existing keys, avoiding token-level indexing and a separate full-length index cache. It uses one indexing query per KV group and shares selected chunks across queries in that group. During prefill, QShare reuses selections across adjacent queries to reduce indexing cost and improve KV reuse and GPU utili
long-contextbenchmark - arxiv:2611.03196 · cs.CLDhārā Bench: An Evidence-Oriented Retrieval Benchmark for Indian Knowledge SystemsAvinash Holenarasipur Nagendra, Anand Kaniyaar Sriranga Prasad, Kousik Sarathy Sridharan, Mohan Raghavan
Dhārā Bench is a retrieval benchmark for Indian Knowledge Systems (IKS), with 325 questions in 12 categories and 60,121 text chunks from 15 source groups. The categories were developed based on 4,500 questions submitted to the Dhārā application; the evaluation set combines corpus-aligned application queries with expert-contributed questions. The annotations comprise 7,578 question-chunk relevance links involving 6,847 distinct corpus chunks. For 34 composite and multi-hop questions, 401 conjugate groups specify alternative sets of passages that jointly support each question. Chunk identifiers
benchmark - arxiv:2611.03183 · cs.AIAuto-Research Agent on MSA Organization to Improve Protein-Protein Complex Structure PredictionZilu Zhang, Jing Bao, Lifeng Qiao, Sheng Xu +3
Protein-complex structure prediction with AlphaFold 3 depends on MSA content, depth, and cross-chain pairing. Current pipelines apply uniform recipes, typically maximizing depth and pairing by species, yet even whether cross-chain pairing helps AF3 remains contested. We show that the value of both depth and pairing is strongly target-dependent, so no single recipe is optimal across interfaces. We introduce an autonomous agent that treats a frozen AF3 as a black-box experimental instrument. For each target, the agent formulates biological hypotheses, analyzes and reorganizes the MSA, and iterat
agentautonomous agentbenchmark - arxiv:2611.03179 · cs.CLEvery Clause Counts: Instruction-Following Reranking by Multiplying Small, Answerable QuestionsS. M. Mahbub Uddin Ahmed, Abdullah Al Musaddiq
People often attach instructions to their search queries, asking for pages written for beginners, documents in a certain language or official sources only. Most rerankers were trained to judge topical relevance and treat such instructions as a few extra words. On the InfoSearch benchmark, even GPT-4o used as a list-wise reranker reaches an average WISE (Weighted Instruction Sensitivity Evaluation) of only 33.5. We present DMReranker (Decision-Model Reranker), a zero-shot reranker that splits an instructed query into a topic and a short list of typed clauses. A small judge model answers one nar
benchmarkjudge model - arxiv:2611.03178 · cs.AICoworkOS:Scaling Agent Coworkers for Real-world AutomationJiajie Jin, Guanting Dong, Chengjun Pan, Xin Deng +36
Rapid advances in cowork agents are shifting human-AI collaboration toward real-world automation, delivering software, designs, and other artifacts. However, such productivity work demands both high-quality delivery on complex productivity tasks and continued monitoring and decision-making over long horizons, a combination that remains an open challenge. We introduce CoworkOS, a general-purpose harness for long-horizon work that separates task strategy from execution management: the model decides actions, delegation, and acceptance, while the runtime maintains state and feedback. Its three-lay
semantic memoryagentpost-traininghuman-in-the-loopbenchmark - arxiv:2611.03171 · cs.LGViAdminShift: A Vietnamese Benchmark for Event-Level Knowledge Updates Built from the 2025 Administrative ReformPhan Binh Nguyen Lam, Nguyen Thai Anh
On July 1, 2025, Vietnam reduced its 63 provinces to 34, abolished its 696 districts, and replaced its 10,035 commune-level units (wards) with 3,321. ViAdminShift, a Vietnamese language resource about this event, consists of 7,995 benchmark items, 2,281 retrieval passages from pre- and post-reform Wikipedia revisions and dated templates, and old-to-new tables that a released script rebuilds from the official conversion table (not redistributed). Changed items store the pre-reform answer (except 121 address items), a $2\times2$ design uses untouched units as controls, and 400 addresses appear w
benchmark - arxiv:2611.03168 · cs.RORollout then Recycle: 3,718 Labelled Real-Robot Episodes from Benchmarking 7 Policies on 12 TasksPraveen Selvaraj, Lorenzo Uttini, Ville Kuosmanen
We release all 3,718 episodes from the ArmnetBench v0.1 evaluation (arXiv:2607.24481), which contains 600 teleoperated demonstrations and 3,118 rollouts of seven imitation-learning and vision-language-action policies on 12 tabletop tasks, recorded on low-cost SO-101 arms; 73% of the rollouts are failures. Every policy for a task was trained or fine-tuned on the same 50 demonstrations and rolled out on the same cell, so the corpus contains many policies' failures on shared tasks. An operator labelled every rollout successful, suboptimal or failure. Each episode has three camera views, joint sta
vision-language-actionbenchmark - arxiv:2611.03166 · cs.LGCausal discovery needs explicit epistemic standardsHarald Kugler, Leonard Henckel, Sebastian Weichwald
There are many causal discovery algorithms, but few guiding principles for choosing between them. Methods are usually compared through three lenses: identifiability, consistency, and benchmarks. Each introduces assumptions about the data-generating process that are often untestable. We argue that there are no shared standards for comparing, prioritizing, or even discussing those assumptions. We support this claim by showing that (i) identifiability assumptions are often preferred by convention rather than by principle, (ii) uniform consistency requires further untestable assumptions, while the
benchmark - arxiv:2611.03163 · cs.ROFineCraft: Enhancing VLA Models for Fine-Grained ManipulationShuaike Zhang, Ziheng Quan, Junhao Ran, Muhe Ding +2
Fine-grained manipulation couples instruction grounding, local visual precision, and sustained execution. We present FineCraft, a framework that adapts pretrained models for fine-grained manipulation in the MetaFine Challenge. A task-routed mixture of policies selects independently adapted DINOv2-DiT or pretrained $π_{0.5}$ policies through a common observation-action interface. Local visual features help identify the instructed object parts, demonstrations and expert continuations teach the required actions, and new observations guide replanning during execution. FineCraft ranks fourth with a
vlavla modelmanipulation - arxiv:2611.03162 · cs.AIBeyond Surface Metrics: Multi-Condition Requirements in Enterprise Maintenance and AI Coding BenchmarksHeeyong Eun
Repository-level AI coding benchmarks measure task success but do not directly capture the conditions required for software maintenance completion. We study multi-condition fulfillment, where a maintenance task must satisfy multiple operational conditions, sometimes with directional dependencies. Using a fixed 18-criterion taxonomy, we analyze 2,925 Apache Fineract issues and 944 tasks from SWE-bench Pro, SWE-PolyBench Java, and SWE-EVO. We examine condition breadth, corpus-level co-occurrence, and source-grounded directional requirement relationships. Among 1,177 benchmark task pairs matched
benchmark - arxiv:2611.03157 · cs.AIToward Knowledge-Centric AI: Discovering, Internalizing, and Using Explicit Knowledge Across Agents and GenerationsYisen Gao, Haoyu Huang, Zhongwei Xie, Yufei Li +6
Knowledge is how understanding accumulates across people and generations: durable records that outlive their producers and that others can check, combine, teach, and correct. AI agents are now built and run by many independent users and organizations, and each agent is upgraded, replaced, or retired within months. Yet what an agent learns stays trapped in model parameters, a transient context window, or a system-specific memory store that is appended to and retrieved from rather than processed. We argue that the next generation of AI should be knowledge-centric, built around an explicit, conti
memoryagentai agent - arxiv:2611.03150 · cs.LGFlowBatt: Flow Matching for Probabilistic Battery Degradation PredictionHamidreza Eivazi, Eslam Sharaawy, Stefan Wittek, André Hebenbrock +2
Battery degradation remains a central challenge in the development and broad application of sustainable energy technologies. Accurate degradation prediction is challenging, as battery aging emerges from complex and heterogeneous interactions among cycling behavior, operating conditions, and cell chemistry. Existing machine-learning approaches typically focus on deterministic end-of-life predictions or modeling degradation curves within fixed chemistry settings, leaving probabilistic state of health (SOH) trajectory modeling across heterogeneous chemistries underexplored. We introduce FlowBatt,
benchmark - arxiv:2611.03146 · cs.LGControlling Sharpening in LLM Post-TrainingSimon Matrenok, Nicolas Menet, Nicolas Flammarion, Andreas Krause
LLM post-training via reinforcement learning has shown consistent improvements in accuracy at the cost of reducing behavioral coverage - a phenomenon known as sharpening that can stall progress on difficult tasks. Repeated sampling objectives such as $\mathrm{pass@k}$ and $\mathrm{max@k}$ are known to mitigate sharpening, but the mechanism behind it remains poorly understood. Moreover, existing approaches for $\mathrm{@k}$ optimization use dedicated algorithms, and no general framework exists for adapting established policy optimization methods from $\mathrm{@}1$ to $\mathrm{@k}$. We derive th
post-training - arxiv:2611.03142 · cs.CLDepthGraft: Structural Regularization through Hierarchical Cross-Layer KV ReconstructionXiaohan Qin, Xiangdong Zhang, Yu Wang, Huaijin Wu +3
Scaling depth is a key route to stronger large language models, but Pre-LN architectures often suffer from depth-wise information dilution, limiting the effective use of deep layers. While cross-layer connectivity has proven effective at mitigating this problem, existing designs usually introduce additional computation or memory access. We revisit cross-layer KV sharing as an efficient form of such connectivity. Beyond cache compression, each KV reuse edge creates a direct gradient path from deep layers to source KV projections, turning KV sharing into an implicit structural regularizer for KV
memory - arxiv:2611.03138 · cs.ROKlaskTron: An Open-Source Platform for Physical Adversarial Multi-Agent RLAswin Ramachandran, Jona Schulz, Maurus Derungs, Carlo Angelini +2
Progress in robot learning, particularly physical multi-agent reinforcement learning (MARL), is currently challenged by a limited availability of accessible, standardized benchmarks. While simulation-based MARL has produced remarkable emergent behaviors, from coordinated team play to complex tool use, translating these advances to physical systems remains difficult. A key barrier is infrastructural: physical platforms are often prohibitively expensive, require specialized facilities, or lack open mechanisms and designs for reproducibility. We introduce KlaskTron, an open-source, low-cost (<\$2
manipulationsim-to-realmulti-agenttool usebenchmark - arxiv:2611.03133 · cs.CLMeet Students Where They Are: Student-Prefix Teacher Adaptation for On-Policy DistillationXue Zhang, Songming Zhang, Yunlong Liang, Fandong Meng +2
On-policy distillation (OPD) mitigates the student's exposure bias by optimizing on student-generated trajectories. However, student-generated prefixes may contain reasoning errors or follow patterns unfamiliar to the teacher, undermining the reliability of its token-level supervision. This degradation becomes more pronounced as prefixes grow longer. To address this teacher-side distribution mismatch, we propose Student-Prefix Teacher Adaptation (SPTA), which adapts the teacher to the states on which it is queried during distillation. Specifically, before OPD starts, we collect student rollout
benchmark - arxiv:2611.03126 · cs.AICUA-Master: Scaling Computer-use Agents through Open-World Recursive ImprovementsXinlei Yu, Gen Li, Ji Zhao, Zhangquan Chen +14
Despite recent advances in CUAs, their development remains constrained by the limited diversity, controllability, and reliability of interaction data. Moreover, conventional optimization pipelines lack iterative feedback, while existing recursive improvement approaches often rely on self-generated experiences, limiting ability expansion and yielding diminishing returns. To address these challenges, we introduce CUA-Master, a nested double-loop recursive improvement framework. At the system level, an outer world-data-model flywheel continuously incorporates open-world corpus and leverages evolv
benchmark - arxiv:2611.03125 · cs.CVSurgPace: Complexity-Aware Progressive Distillation for Surgical Multimodal Large Language ModelsZhicheng He, Xiaohan Xing, Hongqiu Wang, Kun Yuan +6
Multimodal large language models (MLLMs) have shown strong potential in robotic surgery for question answering, scene understanding, and grounded assistance, but their high computational cost limits real clinical deployment. Knowledge distillation provides a practical way to build compact surgical MLLMs, yet direct distillation is not well suited to surgical data because surgical samples vary greatly in difficulty, visual ambiguity, and required clinical context, making uniform teacher-student transfer unstable. In addition, the role of supervision changes during training: early stages require
benchmark - arxiv:2611.03121 · cs.AIComplianceAtlas: A Taxonomy-Driven Agentic AI Framework for Automated AI Act Gap Analysis and RemediationIlias Driouich, Hongliu Cao, Eoin Thomas, Yannick Devaux +1
Assessing regulatory compliance for high-risk AI systems requires systematic identification of technical gaps such as missing controls, in- adequate processes, or shallow evaluation of practices. Compliance officers typically admin- ister structured questionnaires collecting evi- dence of current practices, where organizations respond with descriptions of their existing pro- cesses, technical capabilities, and architectural choices. However, translating these compli- ance questionnaire responses into typed, ac- tionable gap diagnoses remains unsystematic. In this paper, we present a literature
agentagentichuman-in-the-loop - arxiv:2611.03116 · cs.AIFrom Local Skills to Long-Horizon Web Tasks: Progressive Skill Exploration for LLM Web AgentsXueyang Feng, Xiaohe Bo, Xu Chen, Quanyu Dai +1
Large language models (LLMs) are driving autonomous agents toward real-world environments, with the Web emerging as one of the most representative interaction scenarios. However, training Web agents with naive exploration remains challenging: successful trajectories are rare, and the sparse terminal rewards make it difficult for agents to obtain useful feedback from failed trajectories. Inspired by skill acquisition theory, we propose Skill-Composed Progressive Exploration (SCOPE), a short-to-long learning framework for long-horizon Web agents that decomposes complex task exploration into a gr
agentautonomous agentagentic - arxiv:2611.03114 · cs.CV3D Gaussian Splatting within Self-Learned Neural View-Dependent Color FieldsYixiao Yang, Yu-Shen Liu, Zhizhong Han
3D Gaussian Splatting (3DGS) has become a vital tool in novel view synthesis. 3D Gaussian functions with attributes like color, opacity, and scale are stacked together to explicitly model a radiance field. However, discrete 3D Gaussians usually produce poor depth with few geometry details but lots of floaters, which hinders the effective modeling of smooth radiance fields. To address this issue, we propose to learn a radiance field within a continuous view-dependent color field parameterized by neural networks. With continuous constraints in color, the bias on low-frequency signals of neural n
benchmark - arxiv:2611.03103 · cs.ROVLALagBench: Understanding When and Why VLA Inference Latency MattersQi Wen, Borui Xu, Hussein Amro, Hongshi Tan +10
Vision-language-action (VLA) models offer a promising path toward general-purpose robot manipulation. While inference latency can affect task success in a constantly evolving physical world, the relationship between latency and task success rate remains poorly understood. In this paper, we introduce VLALagBench, a comprehensive benchmark for studying VLA inference latency sensitivity across task types, object speeds, execution modes (synchronous and asynchronous), and VLA policies. Using VLALagBench, we find that even a modest inference latency of 100 ms can nearly halve task success on moving
vision-language-actionvlamanipulationbenchmark - arxiv:2611.03101 · cs.LGDiagonal Equivalence in State Space Models: Characterization, Retrieval Rank, and Selective ExtensionsDennis Rochau
We classify real and complex state-space models (SSMs) with linear transitions, distinguishing fixed-basis diagonal equivalence from input-output equivalence and tracking coefficient-conditioning windows under coordinate changes. We characterize diagonal equivalence through commuting transition families whose members are diagonalizable over the chosen field. With time-shared token-local matrix maps, we identify this class's state-mediated kernels with unnormalized scalar-mode linear attention with multiplicative masks. Mamba 3's multiple-input, multiple-output (MIMO) construction lifts one val
memory - arxiv:2611.03100 · cs.AILUCA: A Synthetic World for Experimental ReasoningAnna Gogleva, Thomas Charman, Martin Balla, Karl Gemayel
An experimental agent can succeed for the wrong reasons. Evaluations, however, often focus on the outcome, not the experiments, evidence, and inferences that produced it. We introduce LUCA, a configurable synthetic world for evaluating this process under fully known ground truth. Agents learn unknown regulatory systems through budgeted experimentation, then predict unseen interventions, diagnose hidden faults, and attempt rescues. This setup allows us to evaluate experiment selection, prediction, diagnosis, and intervention separately. Across seven frontier LLMs and three baselines, these capa
agent - arxiv:2611.03097 · cs.CVDR Pixel: Page-Native Deep Research with Vision-Language ModelsTiansheng Hu, Yichuan Wang, James V. Talwar, Nicholas Tomlin +2
Complex documents such as scientific papers contain rich information in tables, figures, and equations; however, existing deep research (DR) agents typically reduce papers to OCR-extracted text. We instead ask whether vision-language models (VLMs) can enable page-native deep research: directly retrieving and reasoning over rendered document page images. We introduce DR Pixel, the first open-source end-to-end multimodal DR agent that uses rendered page images as its sole retrieval and reasoning unit, operating over a corpus of 1.06M pages. To evaluate this setting, we construct DRPixelBench, a
agentagenticbenchmark - arxiv:2611.03094 · cs.AIA Fine-Grained Hierarchy and Benchmark for State Tracking in Language ModelsJackson Petty, William Merrill, Ashish Sabharwal
Theoretical limitations of transformers on state tracking---the ability to track information about entities or variables over long contexts---have motivated new recurrent language modeling (LM) architectures such as Gated DeltaNet (GDN). These architectures have, in particular, been benchmarked against the $S_5$ word problem, a theoretically hard state-tracking task that captures abilities such as code evaluation under arbitrary variable assignment. It, however, remains open whether other, simpler forms of state tracking might both be more relevant in practice and already explain observed diff
long contextbenchmark - arxiv:2611.03093 · cs.LGAn Experimental Study of Hybrid Retrieval, NLI and Open-source LLMs for Scientific Claim VerificationLakshmi Narayanan P, Narenkarthik K, Naren Kumar C
Claim verification is the task of deciding whether a source text supports a claim, contradicts it, or lacks sufficient evidence. Natural Language Inference (NLI) models can judge whether a single sentence supports or contradicts a claim, but they struggle when the evidence is spread across several sentences. Conversely, Large Language Models (LLMs) can reason across sentences, but may draw conclusions the text does not support. In this experimental study, we test whether a conservative fusion rule, labeling a claim Support or Contradict only when both models agree and defaulting to Not Enough
benchmark - arxiv:2611.03092 · cs.AIAvatar Intelligence: Amplifying Human Agency Through Persistent Personal AgentsZheng Wu, Xinbei Ma, Yao Yao, Zhaoyuan Huang +7
Personal AI agents are increasingly capable of handling long-horizon tasks across applications, yet users often need to invest substantial effort in specifying goals, resolving ambiguity, and verifying outcomes. We argue that progress should be measured by how effectively agents amplify human agency: enabling people to pursue their goals with less routine supervision while retaining control over consequential decisions. We advocate Avatar Intelligence, a paradigm of persistent personal agents that represent users' goals, values, and constraints through sustained decision delegation across cont
ai agentagentic - arxiv:2611.03091 · cs.ROKlaskTron-IL: A Contact-Rich Adversarial Imitation Learning Benchmark with a Validated Sim-to-Real BridgeNicklas Sohm, Aswin Ramachandran, Raffaello D'Andrea
KlaskTron is an open-hardware physical platform for two-player adversarial reinforcement learning, with a learned actuator model that supports zero-shot sim-to-real transfer of trained policies. Building on this platform, we contribute KlaskTron-IL: a multi-tier imitation learning dataset (80k human demonstration timesteps collected directly on the physical robot, 320k symmetry-augmented timesteps, and synthetic expert trajectories up to 5M steps generated by the zero-shot-validated PPO policy), and diagnostic baselines that expose three characteristic failure modes of standard IL methods unde
sim-to-realbenchmark - arxiv:2611.03087 · cs.CVMIRA: Reinforcing Multimodal Reasoning via Deceptive Contextual AugmentationZhihan Yin, Jianxin Liang, Yifeng Yao, Nonghai Zhang +4
While Reinforcement Learning with Verifiable Rewards (RLVR) has significantly advanced multimodal reasoning, existing frameworks suffer from a critical "perception-reasoning gap." Due to an over-reliance on seemingly relevant visual tokens, even minor perceptual perturbations can propagate through the reasoning process, leading to compounding hallucinations. To address this issue, we propose MIRA, a training framework that improves robustness to erroneous visual contexts while enhancing logical reasoning ability. By injecting filtered, deceptive visual contexts during training, we construct ch
benchmark - arxiv:2611.03067 · cs.ROPlanContext: Benchmarking In-Context Driving Decision-MakingRuikai Li, Naiyu Gao, Hao Shan, Yizhe Fan +10
Real-world driving requires context-aware decision-making. Existing end-to-end driving benchmarks primarily assess trajectory quality through safety, compliance, comfort, and driving progress, leaving context-dependent decision-making insufficiently characterized. To fill this gap, we present PlanContext, a benchmark for evaluating in-context driving decision-making. PlanContext constructs paired evaluations by varying high-level driving intentions while keeping the observed scene fixed, enabling a natural and controllable assessment of context-dependent decision-making. We further define six
benchmark - arxiv:2611.03066 · cs.AITowards a Trustworthy and Governable Internet of Agentics: An Architectural Perspective on Protocols, Systems, and ServicesHaoyi Xiong, Yi Liu, Jiang Bian, Peng Diao +4
Large language model applications are increasingly being organized as tool-using and multi-agent workflows rather than as isolated chat interfaces. This perspective paper examines the architectural problem that follows from that shift: agents may reason, invoke tools, and delegate work, but dependable deployment still requires interoperability across frameworks, runtimes, data boundaries, and organizational control domains. We study this problem through the lens of the \emph{Internet of Agentics} (IoA), an emerging service-computing view in which heterogeneous agents are treated as governable
agentmulti-agentagentic - arxiv:2611.03063 · cs.AINetworked AI: Architecture, Theory, and ApplicationsLiang Song, Jiayue Jin, Peng Sun, Di Li +8
Recent advances in Artificial Intelligence (AI), particularly large-scale deep learning, have been driven by monolithic or offline training on massive curated datasets. While successful, these approaches face structural limitations: escalating data and computational demands, limited online adaptability, and a lack of system-level intelligence in open, non-stationary environments. This paper presents Networked AI - a networked intelligent system composed of interacting, heterogeneous, and continuously evolving AI agents. We establish a hierarchical layered architecture comprising an Agent Layer
embodiedagentai agent - arxiv:2611.03060 · cs.CLSAGE (Subjectivity-Aware Generation Estimator): A Linguistic Framework for AI-Generated Review Detection *Gonzalo Freijedo Aduna, Charles Yaiche, Patrick Bas, Alda Mari
We propose a simple, linguistically motivated approach to detecting AI-generated peer reviews based on the inherently human trait of subjectivity. We capture subjectivity at two levels: the semantic level of evaluative language and modality and the pragmatic level of speech acts. Using features derived from speech act annotations and model-predicted subjectivity scores, we train a lightweight classifier that achieves an AUROC of 0.9566 with only these linguistic features. Combining its output with the state-of-the-art detector Sem-Detect raises the TPR at 1% FPR of Sem-Detect on the APRDB benc
benchmark - arxiv:2611.03059 · cs.ROBeyondContact: A Reinforcement Learning Framework for Airflow-Based Non-Contact ManipulationPeiyu Yang, Mohamad S. Moudallal, Antonio Langella, Giulio Turrisi +2
Contact-based robotic manipulation is constrained by end-effector reachability and the mechanical properties of objects. Airflow can exert forces without direct robot-object contact, enabling manipulation of objects beyond the reach or when direct contact is undesirable. However, the lack of suitable simulation has hindered the application of reinforcement learning to airflow-based manipulation, limiting the broader adoption of this mechanism compared with contact-based methods. To address this gap, we developed a robot- and task-agnostic airflow extension for Isaac Lab that supports massively
manipulationgripper - arxiv:2611.03055 · cs.AIPRIME: Predictive Routing of Interventions by Modeling Effects for Scientific IdeationFengxian Ji, Xie Yu, Zhexuan Cui, Yuke Li +7
Feedback-driven iteration is widely used to improve LLM-generated scientific ideas, but existing systems either leave revision decisions implicit in generation, align revisions with feedback-localized weaknesses, or restrict action selection to system-specific operations. This is problematic because problems, hypotheses, and solutions are interdependent: the location where a weakness manifests does not itself determine the revision scope that will produce a better successor. We term this phenomenon diagnosis-intervention mismatch. We introduce Predictive Routing of Interventions by Modeling Ef
benchmark - arxiv:2611.03053 · cs.LGOptimal Subgroup Discovery at Every Support ThresholdLincen Yang, Qi Huang, Niki van Stein, Thomas Bäck +2
Subgroup discovery aims to identify regions of the feature space where a target variable deviates from its marginal distribution. A fundamental challenge in this problem is the trade-off between a subgroup's deviation and its support: small subgroups can be highly deviating but statistically meaningless, while large ones cannot deviate much by construction. Existing methods tackle this challenge by collapsing the two into a single scalar quality measure, implicitly committing to one particular trade-off and making it difficult to recover the most deviating subgroup under a user-specified suppo
benchmark - arxiv:2611.03042 · cs.LGBayesian Sequential Optimal Experimental Design for Single-Pixel Imaging using Reinforcement LearningMarcos Antonio Obando, Tianyuan Wang, Tanja Tarvainen, Tristan van Leeuwen +1
Adaptive compressed sensing (CS) aims to select the most informative measurements sequentially, which is crucial for many modern imaging systems. This task can be formulated as a Bayesian sequential optimal experimental design (sOED) problem, where each new acquisition is chosen to maximally reduce uncertainty about the unknown image given prior information and subsequent data provided by acquisitions. However, in single-pixel imaging (SPI), the large design space consists of all possible binary illumination patterns. Directly optimizing the Bayesian sOED objective over this space using conven
agent - arxiv:2611.03039 · cs.LGTSB-SEG: A Systematic Time-Series Segmentation BenchmarkFélix Chavelli, Arik Ermshaus, Fan Yang, Patrick Schäfer +2
Time-series segmentation, studied as either Change Point Detection (CPD) or State Detection (SD), underpins a broad range of monitoring and diagnostic applications. However, the field suffers from a persistent divide between the statistical community (focusing on CPD) and the machine learning and data mining communities (focusing on SD). Consequently, the literature remains fragmented: evaluations typically cover narrow methodological families, datasets are often small or homogeneous, and CPD and SD are rarely evaluated together. We address these gaps with TSB-SEG, a unified benchmark comprisi
benchmark - arxiv:2611.03032 · cs.LGHybrid Federated Learning for Multimodal Fusion under Missing Modalities and Site HeterogeneityXiaofei Wu, Rongmei Liang, Jian Qing Shi
Multimodal fusion is studied under hybrid federation, where data are partitioned across samples and feature blocks. The proposed two-dimensional federated smooth multiple-kernel large-margin learning method (2D-FedSMKL) integrates availability-corrected modality scores, group-kernel regularization, and smooth worst-site risk aggregation. For fixed maps and propensities, the objective is convex and smooth. Curvature bounds and exact synchronous block-gradient equivalence are established for binary and multiclass scores, giving a conditional accelerated rate. An oracle-propensity binary risk bou
benchmark - arxiv:2611.03027 · cs.CVVLM-as-a-Judge for Person Re-IdentificationBessie Dominguez-Dager, Miguel Cazorla, Vincent Lepetit
Cloth-changing person re-identification (CC-ReID) is still very challenging, and no method clearly dominates: methods that lead on one benchmark collapse on the others, because they rest on different identity cues. We show that a general-purpose vision-language model (VLM), used zero-shot as a judge for reranking a short candidate list, beats the expert retrieval models it reranks by a wide and consistent margin. On the four scenarios of CHIRLA, a long-term CC-ReID benchmark that no publicly released Re-ID model is trained on, reranking the candidates of two frozen experts raises rank-1 on eve
benchmark - arxiv:2611.03023 · cs.CVVizDAGuide: Misleading Chart Question Answering in Compact Vision Language Models via DAG-Structured SupervisionHaiyang Zhang, Sicheng Song, Zixin Chen, Jie Gu +3
Misleading charts can lead vision-language models to conclusions that conflict with underlying data. Existing work focuses on inference-time strategies, while training-time supervision remains less explored. We present VizDAGuide, a DAG-structured supervision method for misleading-chart question answering. The method represents intermediate reasoning states and their dependencies as a directed acyclic graph, assigns supervision weights according to their roles, and serializes the graph into a unified training target for open-ended and multiple-choice QA with single-pass inference. Experiments
benchmark - arxiv:2611.03012 · eess.SYLow-rank corrections to a shared reference Jacobian for GPU-parallel AC $N\!-\!1$ contingency analysisMarc Hunkemöller, Nico Westerbeck, Joost van Dijk, Lars Schewe +1
Contingency analysis repeatedly solves the power-flow equations under many single-element outages, an increasingly demanding workload as operating margins tighten and renewable variability grows. Graphics processing units (GPUs) are attractive for this task, but a change in topology alters the sparsity pattern of the Jacobian, preventing the uniform batched factorization on which efficient GPU solvers rely. This paper presents a GPU-parallel Newton--Raphson solver for $N\!-\!1$ contingency analysis that separates the change in operating state from the change in topology. A single reference Jac
benchmark - arxiv:2611.03005 · cs.AIBLPO: Beyond-Local Policy Optimization for Long-Horizon LLM AgentsXuanren Chen, Chongyang Tao, Wenpeng Hu, Yong Wang +3
Long-horizon tasks require large language model (LLM) agents to coordinate interdependent actions, a capability central to automating complex workflows. Stepwise group-based reinforcement learning provides finer credit than trajectory-level supervision, but reliable comparisons must account for differences in interaction history. Enforcing this consistency within the limited rollouts of each task can leave insufficient support for estimating action preferences. We propose Beyond-Local Policy Optimization (BLPO), which estimates task-progress advantages from experience shared across tasks and t
memoryllm agent - arxiv:2611.02995 · cs.CVILLRAW: Illumination-Aware RAW Object DetectionXinzhe Wang, Xiao Liang, Xiaoli Zhang, Kang Ma
Existing RAW-based detectors typically introduce learnable ISP modules to convert RAW inputs into sRGB-like formats to fit off-the-shelf detectors. However, this RAW-to-sRGB paradigm decouples representation learning from the downstream detection task. Crucially, it forces the network to inherit human-centric visual priors while discarding intrinsic RAW advantages, such as linear sensor responses and high-bit-depth, minimally processed scene cues. In this paper, we propose \textbf{ILLRAW}, an \textbf{ILL}umination-aware end-to-end \textbf{RAW} object detection framework. Instead of reconstruct
benchmark - arxiv:2611.02978 · cs.AICAREFUL-Bench: Effect-Aware, Judge-Free Evaluation of Multi-Turn LLM Function CallingAlexander Kovalenko, Artem Redchych
The multi-turn checker of the Berkeley Function Calling Leaderboard (BFCL) evaluates final state correctness but ignores the effects of intermediate calls: extra calls are free, function documentation does not describe side effects, and the scripted user never responds to proposals. Consequently, models asking for confirmation before destructive actions are penalized; 260 of 734 reference turns in the base category perform destructive actions without confirmation. We make BFCL effect-aware without replacing it. We infer effect declarations for all 128 functions by executing them in BFCL simula
benchmarkleaderboard - arxiv:2611.02968 · cs.ROReContact: Reconstruct and Retarget Human Contacts from Video for Contact-Rich Humanoid ManipulationZhongkai Gu, Yang Liu, Wei Zhu, Mitsuhiro Hayashibe
Contact transitions allow humanoid robots to redistribute object support across the whole body, enabling manipulation beyond hand-only grasps. Human videos provide a scalable source of such transitions for robots to learn from. However, body occlusion disrupts object pose tracking and corrupts retargeted motion. Moreover, accurately reproducing the object trajectory does not ensure that the robot applies stabilizing forces at the intended contacts, risking slips or drops in later phases. To address these challenges, we propose ReContact, which reconstructs and retargets human-object interactio
manipulationhumanoidgrasp - arxiv:2611.02950 · cs.LGAction Verification through Conjecture and Refutation for Language Model AgentsMinhyeon Oh, Chiyeong Heo, Geonwoo Noh, Hoyoung Kim +2
Language model agents increasingly act on users' behalf by manipulating external environments through tools, often from high-level instructions that leave critical task- and environment-specific context unstated. To prevent resulting misaligned actions, recent work verifies proposed actions before execution, largely by establishing whether the conditions supporting them hold. We argue that this confirmation-oriented view leaves a fundamental blind spot: a wrong action often passes every check that asks whether it is right. This motivates a different question for verification: rather than askin
agent - arxiv:2611.02947 · cs.ROPositive Safety Margins from Artificial Potential Field Reward Augmentation for Size-Robust Obstacle Avoidance under Partial ObservabilityZiyi Hao
Collision-free reaching is central to manipulation, yet real cells often leave the size and exact position of an obstacle unobserved, which excludes reinforcement learning controllers that take obstacle geometry as input. A reward limited to a collision penalty does not depend on the clearance outside the obstacle, so under a step cost its learned policy skims the obstacle surface and fails once the obstacle radius exceeds its training value. We investigate whether a single analytic reward term can induce a safety margin without exposing the obstacle to the policy. We augment the reward of a s
manipulationfranka - arxiv:2611.02940 · cs.AIPlagiaCorpus: A Bilingual, Adversarial, and Cross-Lingual Dataset for Forensic Plagiarism DetectionAdel Ammar, Serry Sibaee, Wadii Boulila
PlagiaCorpus is a bilingual Arabic-English dataset for benchmarking plagiarism and text-reuse detection, including reuse produced by large language models (LLMs) and reuse across languages. It contains 198,643 text records: a source corpus of 34,885 passages drawn from English and Arabic Wikipedia and two openly licensed book collections, and 163,758 queries (147,381 for training and 16,377 for testing). The queries cover five reuse classes of increasing difficulty, namely verbatim copying, character-level obfuscation, same-language paraphrasing by two different LLMs, and English--Arabic trans
benchmark - arxiv:2611.02937 · cs.RORESONATE: a Real Social Navigation Testbed for Autonomous Service RobotsStefano Trepella, Andrea Ostuni, Mauro Martini, Irene Borgini +1
Social navigation is essential for autonomous service robots operating in human environments, but progress remains constrained by the lack of extensive real-world benchmarks. This paper presents the Real Social Navigation Testbed (RESONATE), a reproducible benchmark for human-aware navigation based on a unified robotic platform, a shared ROS 2 software stack, precise trajectory ground truth, and structured social interaction scenarios. We evaluate eight controller configurations, combining classical and predictive local planners with different social modeling strategies, over eight representat
benchmarkevaluation protocol - arxiv:2611.02936 · cs.LGWhy Tabular Foundation Models Excel on Relational Data They Have Never SeenLinjie Xu, Yiqun Hu, Jixiang He, David Wipf
Given a relational database (RDB) storing heterogeneous tabular information, how can we predict missing (or future) values in some target column of interest? As candidate targets are numerous in enterprise settings, it is preferable to avoid learning a new model from scratch each time there is a new prediction task. Frozen foundation models (FMs) based on in-context learning (ICL) and RDB-specific pre-training provide a viable solution, but the ideal design still remains largely unresolved. Perhaps unexpectedly, preliminary evidence has recently shown that existing single-table FMs (aka TFMs),
agentic - arxiv:2611.02935 · physics.opticsElectromagnetic Information Theory for Programmable MetasurfacesPhilipp del Hougne, Marco Di Renzo, Owen D. Miller, David A. B. Miller
Programmable metasurfaces enable fine-grained control of electromagnetic waves and are creating new opportunities for information extraction, information transfer and information processing. Yet, despite exciting experimental demonstrations, their fundamental capabilities remain poorly understood. In this Perspective, we argue that programmable metasurfaces, and programmable wave systems more generally, call for an electromagnetic information theory that addresses limits at two levels, in a unified manner across applications like communications, sensing and computing. First, for a given physic
manipulation - arxiv:2611.02932 · cs.LGILR-SMO: Iterative Latent Refinement for Robust Spatial Multi-Omics IntegrationAnqi Yu, Xudong Xu, Jianzhi Lu, Yuqi Sun +4
Spatial multi-omics technologies jointly profile diverse molecular modalities with spatial context, providing a comprehensive view of cellular heterogeneity and tissue organization. To integrate spatial multi-omics data and identify spatial domains, a wide range of unsupervised methods has been proposed. However, recent approaches rely on single-step fusion, directly aggregating heterogeneous modalities into a shared representation, making embeddings sensitive to modality imbalance and measurement noise that are inherent to spatial multi-omics data, as well as unstable optimization under weak
benchmark - arxiv:2611.02931 · cs.CVTraining-Time Domain Transfer for Generator-Free Diagnostics on Novel Modalities in Computational PathologyJulia A. Mielcarz, Andrea Camilloni, Nita Mulliqi, Jessica A. Frank +4
Novel imaging modalities promise improved diagnostics, but AI development is constrained by scarce training data. Inference-time domain transfer can address this by mapping novel modalities to established ones, enabling reuse of existing diagnostic models. However, this introduces hallucination risks, computational overhead, and regulatory complexity. Inspired by sim-to-real transfer, we propose training-time domain transfer, generating synthetic target-domain training data from archived routine samples to enable direct diagnosis on real inputs without inference-time generation. Using unstaine
sim-to-realbenchmark - arxiv:2611.02930 · cs.ROPointPilot: Unified Robotic Manipulation via Object-Centric 3D Point Flow PredictionYihan Fang, Yuxin Tian, Ruixiang Wang, Zhanqi Zhang +6
Generalization remains a central challenge in robotics, calling for policies that address domain variations from multiple dimensions such as diverse embodiments, camera poses, and/or scene layouts. In order to approach this problem, we introduce PointPilot, a task-focused and 3D aligned manipulation framework built on a 3D space point-flow prediction. Rather than modeling the global scene, PointPilot only represents task-focused objects (robot grippers treated as objects), avoiding the prediction of potentially redundant scene parts unrelated to the manipulation tasks. Moreover, PointPilot ali
manipulationaction headgripper - arxiv:2611.02929 · cs.CVComprehensive Evaluation of Monocular and Multi-view 3D Human Pose Estimation Models for Fitness Exercise AnalysisNéstor Ortega, Agustín Salgado, Javier Sánchez
Three-dimensional human pose estimation has become a fundamental component of computer vision systems for sports analytics, rehabilitation, and motion understanding. Although recent deep learning approaches have achieved remarkable accuracy on standard benchmarks, their performance under realistic sports scenarios remains insufficiently studied. Furthermore, the trade-offs between recent monocular and multi-view approaches, as well as the benefits of post-processing techniques for improving temporal consistency, have not been systematically evaluated on fitness-oriented datasets. This paper ev
benchmarkevaluation protocol - arxiv:2611.02922 · cs.ROAn intelligent satellite must do nothing wellFabrizio Pilone, Giordana Bucchioni
Autonomous inspection of a resident space object is usually posed as a contest between classical next-best-view planning and reinforcement learning. We argue that the choice of algorithm is the least informative of the design axes. An inspector spacecraft has no rest state: when it stops thrusting, orbital dynamics carry it along a natural trajectory that can be used to observe the target at no fuel cost. On a controlled ladder in which one design axis is varied at a time, fuel efficiency turns out to be set by three things: whether the agent is allowed to drift, whether its state records wher
world modelagent - arxiv:2611.02912 · cs.CLA State Change Is Local, Its Consequences Are Not: Dependency-Aware State Revision for Long-Term Agent MemoryYiqi Wang, Tongyuan Zhao, Jiaqi Zhang, Yiqun Duan +3
Long-lived agents rely on persistent memory of users, tasks and decisions, and this memory must stay consistent with a world that keeps changing. When an upstream fact changes, however, memory systems can revise the contradicted memory itself while the beliefs, plans and actions derived from it remain accepted, even though they have lost their support. We therefore formulate dependency-aware state revision, which asks not only what changed but also which dependent states should no longer be trusted. To address it, StateGraph maintains an evolving state dependency graph whose edges are recovere
memorypersistent memoryagent memoryagentbenchmark - arxiv:2611.02909 · cs.MACloudra: Geometric Reasoning and Scientific Validation for Autonomous Reaction Mechanism ExplorationChengyi Yang, Qianxin Zhou, Yang Li
Autonomous reaction-mechanism exploration requires proposing reaction steps, preparing three-dimensional structures, and determining which reactant and product a computed saddle point connects. Large language model (LLM) agents can misjudge spatial relationships and confuse a computed structure with the species or step they intended to study. We present Cloudra, a multi-agent framework that constrains mechanistic output through a Delivery DAG of chemical objects. Species nodes carry geometries and bonding topology; elementary-step edges carry transition states and reaction connections between
multi-agentagent framework - arxiv:2611.02905 · cs.AIWhat makes a theorem interesting? Operationalizing interestingness for open-ended explorationSilvère Gangloff
An agent that explores, whether it generates conjectures or research ideas, must decide which of them are worth keeping. This judgment of interestingness is usually delegated to a hand-picked proxy or to a language model, neither of which is known to correspond to what humans actually value. I study it for mathematical theorems, by observing which results mathematicians single out in formal libraries and in papers. I model each way of singling out results as an evaluator that weights features of a result's statement, its proof and dependencies, and split its weights into a shared and a subject
memoryagentevaluator - arxiv:2611.02903 · cs.AIMaswao: An automatically collected, prepared, and benchmarked dataset for the South African Sign languageThapelo Sindane, Unarine Netshifhefhe, Vukosi Marivate, Abiodun Modupe +2
Recent advancements in sign language processing can be attributed to developments in both large-scale dataset collection and algorithmic complexity in areas such as pose estimation techniques, signal processing through Transformer models, and model optimization. However, only a handful of sign languages, such as the American, British, and German, from over 200 globally recognized sign languages, reap the benefits of such advancements, while the rest lag behind. In this study, we aim to address this technological disparity by proposing a framework for the automatic collection and preparation of
benchmark - arxiv:2611.02902 · cs.CVTailorRig: Controllable and Structured Auxiliary Skeleton Generation for Garment-Aware RiggingYajing Chen, Xin Chen, Saidan Lu, Yeqi Yuan +4
Humanoid rigging for modern characters with complex garments remains challenging due to the need for dense auxiliary skeletons and precise skinning control. Existing automatic methods primarily recover rigs from geometry, providing limited artist control over auxiliary articulation and often lacking explicit semantic or kinematic-chain structure for dense garment bones, leading to instability and weight bleeding. We present TailorRig, a framework that reformulates garment-aware humanoid rigging as a structured, controllable full-rig generation problem. This formulation shifts the goal from aut
humanoid - arxiv:2611.02897 · cs.CVPhotoGuider: Composition-Aware Cropping and 360$^{\circ}$ View Framing with Explainable Photographic AssessmentsYuan-Kang Lee, Kuan-Lin Chen, Chia-Che Chang, Yu-Lun Liu
Traditional image cropping assumes the photograph already contains the composition worth keeping, and asks only where to crop. A full 360$^{\circ}$ scene offers no such guarantee: the viewing direction in a scene has to be discovered before anything can be framed. We present PhotoGuider, a measure-then-frame method that answers both questions with compact learned scorers over one shared set of composition-aware features. The designed features are extracted using foundation and task-specific vision models and structured into four photographic tokens: Subject, Composition, Aesthetics, and Integr
benchmark - arxiv:2611.02889 · cs.AITug of Watts: A study on the Energy - Accuracy Trade-off of Large Language Models on Medical Question Answering TasksViju Sudhi, Giulio Billi, Andreea Scrob, Regina Stodden +1
Large Language Models (LLMs) are increasingly used in medical question answering (QA) systems. While recent works aim at incrementally improving the downstream task accuracy, an empirical estimation of the models' energy consumption and the associated carbon footprints is often overlooked. In this study, we demonstrate how practitioners can estimate energy consumption and arrive at reliable emission estimates for medical QA tasks available in the OpenMedical-LLM leaderboard by investigating 13 LLMs from 3 model families ranging from 270M to 14B parameters. We discuss how one can critically int
leaderboard - arxiv:2611.02888 · cs.LGA Universal Policy Wrapper With GuaranteesGeorgiy Malaniya, Anton Bolychev, Pavel Osinenko
Many control tasks require driving the system into a prescribed goal set and subsequently keeping it there. The considered setting assumes that a working controller is available that provably reaches and maintains the goal, albeit with suboptimal performance. This controller is referred to as the fallback policy. Conversely, although state-of-the-art reinforcement-learning (RL) algorithms can discover policies with impressive empirical performance, the neural networks at their core act as opaque black boxes learned through data-driven exploration and, as such, provide no theoretical mechanism
online learning - arxiv:2611.02883 · cs.AIIntegrating Domain Adaptation and Retrieval-Augmented Generation for Saline-Alkali Land Reclamation PlanningZuguang He, Congcong Zheng, Wanyue Chen, Wenya Zhou +4
Reclamation of saline-alkali land requires management plans that account for soil and water conditions, treatment evidence, and financial constraints. This study develops Wenyao, an agricultural decision-support system integrating domain adaptation with retrieval-augmented generation. A bilingual corpus of 8,664 research papers was processed using a 110-indicator schema to construct SA-Bench, comprising 4,860 paper-level planning instances. A QwQ-32B model was adapted through low-rank supervised fine-tuning and combined with case retrieval for budget-conditioned planning. Evidence extraction a
retrieval-augmented - arxiv:2611.02877 · cs.CVProtoFusion: Protocol-Guided Prompt Refinement for Generative Infrared and Visible Image FusionPengwei Liang, Junjun Jiang, Junhui Hou
Infrared and visible image fusion integrates complementary information from both modalities to provide a more comprehensive representation of the scene. We propose ProtoFusion, a protocol-guided generative framework for infrared and visible image fusion. Instead of relying on fixed fusion operators, hand-crafted objectives, or manually specified one-shot prompts, ProtoFusion formulates fusion as protocol-constrained generation. It represents fusion intent with a structured protocol, consisting of environment state cards, grounding principles, and a style profile, which explicitly specifies whi
benchmark - arxiv:2611.02876 · cs.LGIdentifiable Signed Causal Learning via Latent Stochastic Differential Equations for Inferring Brain Dynamic Effective ConnectivityYiding Wang, Chen Qiao
Understanding how brain regions interact to support cognitive tasks requires characterizing directed causal influences among them, known as effective connectivity (EC). As inter-regional coupling changes over time, a dynamic EC (dEC) estimate is required to track this temporal variation. Existing dEC learning methods typically lack a causal rationality guarantee or an identifiability analysis, and output unsigned connectivity that cannot distinguish excitatory from inhibitory influences, which limits further study of brain mechanisms. We propose ISCL, a latent stochastic differential equation
latent dynamicsbenchmark - arxiv:2611.02875 · cs.LGAGiR: Mitigating Gift Over-Reliance in Mixed-Motive GamesWoohyeon Byeon, Seongmin Kim, Jiwon Jeon, Woojun Kim +1
In this paper, we consider mixed-motive games (MMGs) which model various real-world multi-agent scenarios where self-interested agents must both cooperate and compete. We identify an issue with existing gifting methods, where agents become overly dependent on incoming gifts and fail to learn reward-yielding behaviors, a phenomenon we call gift over-reliance. To address this, we propose a refusal-augmented gifting framework that balances cooperation and competition in MMGs. Our approach introduces an adaptive refusal mechanism that allows agents to selectively reject a portion of received gifts
multi-agent - arxiv:2611.02871 · cs.ROReGATE: Recurrent Gated Visual Fusion for Reactive Humanoid-Object InteractionHao Zhu, Jorge Peña-Queralta
Humanoid loco-manipulation requires coordinating object acquisition, transport, and release from onboard sensing while responding to user commands and incomplete visual observations. We present ReGATE, a whole-body controller that combines egocentric depth, proprioception, and online locomotion and contact commands. Built on a teacher-student framework, ReGATE integrates interaction-state estimation through a recurrent adapter that predicts object geometry and physical contact without an external object-pose estimator. Head-wise and global gates regulate visual updates to a proprioceptive path
manipulationhumanoidwhole-body controlgrasp - arxiv:2611.02866 · cs.LGTensorLOB: Structure-Preserving Tensor Learning for Limit Order Book ForecastingYuxuan Gu, Danilo Mandic
The limit order book (LOB) records outstanding buy and sell orders across multiple price levels. Its state evolves over time across market side, price level, and order attribute, naturally forming a multiway structure. However, existing LOB models typically flatten these dimensions into a matrix model, thus obscuring the existing structural relationships. Starting from first principles and the market-microstructure intuition of the LOB, we design TensorLOB to operate entirely in the tensor domain. TensorLOB uses mode-wise tensor contractions for feature extraction and directly contracts the le
benchmark - arxiv:2611.02856 · cs.LGKolmogorov-Arnold Networks vs. Multi-Layer Perceptrons: A Rigorous Statistical Benchmark on Clinical Tabular DataRafał Wodziński, Paweł Szczerbiak
Kolmogorov-Arnold Networks (KANs) have recently been proposed as a promising alternative to Multi-Layer Perceptrons (MLPs). However, empirical evaluations on structured medical datasets remain scarce, and commonly reported metrics like AUROC can yield overly optimistic performance estimates on imbalanced clinical cohorts. In this study, we present a standardized benchmark comparing eleven architectures - nine KAN variants (FastKAN, WavKAN, ReLUKAN, ChebyKAN, TaylorKAN, GramKAN, JacobiKAN, LegendreKAN, and TabKAN), a tuned StandardMLP, and a Random Forest baseline - across seven heterogeneous m
benchmark - arxiv:2611.02853 · cs.CVScaling Action-Free Video Latent Prediction for Transferable Representations in End-to-End Autonomous DrivingXingtai Gui, Xubin Zhou, Yucheng Zhou, Dongqian Guo +3
End-to-end autonomous driving policies require effective visual representations that capture both spatial scene structure and temporal dynamics. Video latent prediction has emerged as an effective paradigm for learning spatiotemporal visual representations, yet it remains unclear whether scaling this objective with action-free driving videos can yield representations that benefit end-to-end planning. We study action-free video latent prediction as a scalable pretraining objective and introduce ScaleDrive, a driving-oriented framework pretrained exclusively on large-scale internet open-world dr
benchmark - arxiv:2611.02849 · cs.AITLAPS-Bench: Formally Proving Correctness of Real-World System Specifications with TLA+ Proof SystemQian Cheng, Ruize Tang, Saad Mohammad Rafid Pial, Zhizhen Cathy Cai +12
TLAPS-Bench is a benchmark and execution harness for constructing machine-checked proofs for real-world system specifications in TLA+ with agentic AI. Proving correctness of complex system specifications requires discovering inductive invariants, reasoning about intricate state transitions, and decomposing proofs into tractable obligations. TLAPS-Bench provides a proof harness that supports iterative proof development atop the TLA+ Proof System (TLAPS), where agents are equipped with proof tools and libraries, with safeguards against reward hacking. Using the harness, we have employed AI agent
ai agentagenticbenchmark - arxiv:2611.02846 · cs.AISpecOPD: Branch-Parallel On-Policy Distillation for Speculative DecodingTianyu Liu, Jun Zhang, Wenjie Huang, Yuhao Shen +11
Speculative decoding (SD) accelerates large language model inference by reducing the sequential cost of token-by-token generation. Advanced SD methods reduce proposal cost with parallel drafters that predict entire blocks in one forward pass, but overall speedup still depends on acceptance length. On-policy distillation (OPD) can push well-trained drafters beyond acceptance-length plateaus, yet conventional pipelines collect supervision through sequential draft-and-verify rollouts. Our key insight is that on-policy supervision need not inherit these dependencies: a draft block can be supervise
agentic - arxiv:2611.02843 · cs.AIMath Foundation Formalization Agent: A General-Purpose Approach to Formalizing Mathematical FoundationsJiaxin Xie, Guowei Dou, Xiang Ji, Wensheng Yu
Mathematical formalization plays an important role in establishing rigorous and machine-verifiable foundations for mathematical reasoning. With the development of large language models (LLMs) and interactive theorem provers, intelligent agents offer a promising approach to automating the formalization of mathematical theories. This paper presents the Math Foundation Formalization Agent (MFF-Agent), a general-purpose intelligent agent designed to support formalization tasks across multiple mathematical domains. The agent integrates the reasoning capabilities of large language models with the fo
agent - arxiv:2611.02842 · cs.LGNIR: Approximate Neural Instance Retrieval for ALCIQN'Dah Jean Kouagou, Pritilata Saha, Caglar Demir, Axel-Cyrille Ngonga Ngomo
Instance retrieval is the main bottleneck of concept-learning systems based on refinement operators. We present NIR, a family of neural instance retrievers for ALCIQ that output graded membership scores from knowledge graph embeddings. We compare tokenization-based encoders (GRU, LSTM, Transformer) with a Compositional encoder that mirrors DL operators via primitive sub-networks. On eight benchmarks, all variants approximate Pellet closely on complete knowledge bases, while the Transformer model is most robust under incompleteness. The source code of our approach is at https://github.com/dice-
knowledge graphbenchmark - arxiv:2611.02839 · cs.AICARE: Grounded Recursive Self-Improvement for Production Customer-Service AgentsYanbei Jiang, Chao Lei, Rena Gao, Xiaochen Zhao +16
Recursive self-improvement (RSI) in production agents involves transforming accumulated experience into reliable model updates while adapting to evolving business policies. We present CARE (Customer-service Agent Retrospective and Evolve), a grounded RSI framework that operationalizes this process in customer service. Its goal is to replace a frontier proprietary incumbent with an in-house mid-sized model that continues learning after deployment. After a one-time cold start on governed production traces, the current model serves production traffic alongside the incumbent, and its own failures
agentself-improvement - arxiv:2611.02835 · cs.CLOmni-Life: Benchmarking Omnimodal Real-Time AI Assistants in People's Daily LivesMengze Hong, Ziwei Shang, Di Liang, Zijian Li +8
Real-time omnimodal assistants are expected to support people in everyday life, but audiovisual understanding alone does not establish whether they can provide timely and appropriate help. We introduce Omni-Life, a benchmark grounded in real-world user needs, deployment scenarios, and device-dependent observations. It combines controlled video synthesis for hazardous situations and edge cases with annotated real-world videos under a shared temporal protocol. The benchmark contains 400 videos: 290 requiring assistance, with 555 required intervention events, and 110 requiring silence throughout
benchmark - arxiv:2611.02833 · cs.LGReimagining computational quantum research with AI scientistsYuxuan Du, Hanqing Zhao, Xinbiao Wang, Yiming Lu +2
AI scientists are expanding automated research across scientific domains, yet their ability to turn the scientific record into a foundation for autonomous discovery remains elusive. Here we uncover a fundamental barrier to this transition and demonstrate a route to overcome it in computational quantum science, where advances shape both what we can learn from complex quantum systems and what quantum computing can ultimately deliver. To identify this barrier, we introduce OmniQ, an expert-audited virtual research environment that converts published studies into executable reproduction tasks. Acr
agentautonomous agentagenticself-improving - arxiv:2611.02830 · cs.AIThink with Skills: Pseudocode Planning Before Execution for Advanced ReasonersBoshen Zhou, Boxuan Yue, Xiang Li, Shanchao Yang
Language model agents often struggle with complex tasks even when equipped with expert-written skills. Such skills encode declarative knowledge, whereas successful execution requires procedural knowledge tailored to the task at hand. Natural language skills may leave conditions and dependencies implicit, and tasks involving multiple skills further require agents to coordinate intermediate results and execution order. We introduce Think with Skills, a method that trains reasoning models to turn relevant skill descriptions into task-specific pseudocode plans before execution. The pseudocode expl
agent - arxiv:2611.02826 · cs.AIRxnMiner: A Multimodal Agent for Structured Reaction Data Extraction from Chemistry FiguresJinbo Hu, Songren Zhang, Chuanxun Fu, Jinrui Huang +18
Reaction data underpin machine learning for reaction prediction, condition recommendation, and retrosynthetic planning, yet extracting these data from literature figures still requires substantial manual effort. Many figures encode multiple experiments through shared templates, substituent tables, and product grids. Recovering concrete reaction records therefore requires molecular recognition, association of evidence with individual experimental entries, and reconstruction of indirectly specified participants. We present RxnMiner, a multimodal agent system that combines an expert-informed extr
agentagent systembenchmark - arxiv:2611.02823 · cs.LGEfficient Detailed Scheduling of Multiproduct Pipelines using Action-Decoupled Deep Reinforcement LearningKai Shao, Ling Jian, Liming Liu, Serguey V. Bogomolov +1
Multiproduct pipelines transport refined products to delivery stations in sequential batches, and their detailed scheduling must meet market demand while respecting nonlinear hydraulic constraints. Retaining the full hydraulic nonlinearity yields a mixed-integer nonlinear program that is computationally intractable, while the linearization commonly adopted for solver-based optimization degrades hydraulic accuracy without resolving the heavy computational burden. This paper develops an action-decoupled proximal policy optimization (PPO) framework that efficiently produces detailed schedules whi
benchmark - arxiv:2611.02822 · cs.LGDisen-Forcing: Disentangling Semantic Anchoring from Motion for Autoregressive Video DiffusionZhengyang Yu, Akio Hayakawa, Masato Ishii, Qingtao Yu +4
Autoregressive video diffusion models (AR-VDMs) enable real-time, long-horizon video generation but degrade rapidly once the inference horizon outruns the training horizon, due to compounding exposure bias. To stabilize long rollouts, recent methods heuristically adopt frame sink, keeping initial frames as a persistent model context, inspired by attention sinks in LLMs. We revisit this design and find that the mechanism by which initial frames stabilize long-range consistency in AR-VDMs is not an LLM-style attention sink; instead, long-range consistency is anchored through a sparse subset of a
benchmark - arxiv:2611.02816 · cs.AIDiscovering hierarchical governing equations for community-structured complex network dynamics with LLM agentsJinlei Zhang, Zixuan Zhang, Hao Qiu, Feng Chen +7
Governing equations provide an interpretable description of the mechanisms underlying complex network dynamics, yet their discovery from observational data remains challenging when shared community-level dynamics coexist with individual-level variations and the relevant functional structures may be absent from the initial candidate library or not explicitly specified in advance. Here we formulate this task as a hierarchical law-discovery problem and introduce SAGE, a large language model-guided self-reflective agent framework for discovering community-shared governing laws and individual-speci
agentllm agentagent framework - arxiv:2611.02815 · cs.AIPredictive Safety Guardrails Do Not Always Make LLM Agents SaferWeixiang Zhao, Zhihao Liu, Guancheng Li, Yanyan Zhao +2
Large language model agents increasingly take consequential actions in digital and physical environments, where safety risks may emerge only after multiple interaction steps. This has motivated world-model-driven predictive safety guardrails, which anticipate future consequences before execution. However, existing evaluations largely focus on end-to-end safety outcomes, leaving unclear whether safety gains arise from reliable prediction and faithful use of predicted risks. In this work, we systematically diagnose the prediction-to-action pipeline along three dimensions: how world-model predict
llm agent - arxiv:2611.02810 · cs.CLTRIO: Task Refinement with Individual Optimization for Retrieval-Augmented GenerationYingtao Ren, Xiao Luo, Yiwei Fu, Jinzhao Zhou +2
This paper studies retrieval-augmented generation (RAG), which leverages external knowledge to improve the performance of language models. Despite the remarkable progress, current RAG approaches are still far from satisfactory due to inadequate retrieval and potential hallucinations. Towards this end, we propose a novel approach named Task Refinement with Individual Optimization (TRIO) for RAG. The core idea of our TRIO is to equip three well-designed agents with an iterative refinement paradigm by aligning their individual optimization objectives. Specifically, we first expand ambiguous queri
retrieval-augmentedragagentcurriculum learningiterative refinementbenchmark - arxiv:2611.02809 · cs.AIOffline Approximate Value Iteration: A Reinforcement Learning Approach for Satellite Constellation DeploymentTristan Leuridan, Yiping Fang, Adam Abdin
This paper introduces a reinforcement learning-based approximate dynamic programming approach to optimize the deployment of Walker-type satellite constellations under uncertainties, including launch failures and satellite malfunctions. We formulate the problem as a Markov Decision Process that incorporates low-injection orbit launches, continuous low-thrust orbit raising, and variable satellite loads. We adopt the offline approximate value iteration solution method, using linear function approximation with basis functions and action pruning strategies to substantially reduce the computational
memory - arxiv:2611.02807 · cs.AICreating Hybrid Interactive Narrative Systems with LLM: From Formal Description to EvaluationJules Clerc, Domitile Lourdeaux, Mohamed Sallak, Marc Ravaine
Interactive narratives let users shape a story while an Experience Manager (EM) keeps it coherent, balancing player freedom against authorial control. Large Language Models now enable a new generation of hybrid EMs, which combine run-time generation from free-text input with symbolic narrative control. However, their heterogeneity makes them hard to compare: A manager's behaviour is generally tied to the unique and specific scenario in which it operates, and much of its decision-making takes place outside of any explicit and verifiable structure. We address this with an LLM-based narrative pla
world model - arxiv:2611.02797 · cs.LGSelfImprove-Bench: Evaluating Generalization in Self-Improving AgentsBangrui Xu, Jie Yang, Sidi Chen, Xuanhe Zhou +5
Recursively self-improving (RSI) systems are increasingly viewed as a promising means of accelerating progress toward artificial general intelligence. They aim to autonomously improve their capabilities by refining their working pipelines and recursively improving the strategies that guide subsequent refinements, with gains generalizing to unseen scenarios.However, existing evaluations lack reliable metrics to distinguish capability gains from scenario fitting, sustainable and extensible protocols for unseen tasks and difficulties, and fine-grained attribution of improvement to evolution trace
self-improvingevaluation framework - arxiv:2611.02794 · cs.ROSomaAlign: Reliability-Aware Multimodal Representation Learning for Soft Robotic GraspingHaoyuan Wang, Shuai Peng, Yongkai Liao, Yuxiao Sun +3
Robust multimodal perception for soft robotic grasping is challenged by modality-dependent noise, semantic ambiguity, and limited coordinated multisensory data. We introduce Soma, a dataset containing 100 objects, 1,000 continuous grasping trajectories, and 74,914 multimodal samples grouped by software acquisition cycle, spanning remote RGB, hand RGB, depth, tactile, IMU, and structured text observations. We also collect Soma-Success, an independent grasp-outcome dataset with 1,000 trials (998 retained after quality control), to evaluate frozen-encoder transfer for post-hoc grasp outcome class
manipulationtactilegrasp - arxiv:2611.02791 · cs.CLPGuard: A Native, Streaming, and Plug-in Guardrail for Multimodal Large Language ModelsJunyu Lin, Mengjie Wu, Yonghao Dong, Yuchuan Fu +10
As streaming generation becomes the dominant serving paradigm for both text-only and multimodal large language models, ensuring safety throughout the generation process has become a critical challenge. We introduce PGuard, a native, streaming, and plug-in guardrail that transforms safety monitoring from an external serving dependency into an integrated capability of the backbone model. PGuard directly leverages the backbone's intermediate representations for token-level safety detection, without altering its original inference pathway or generation behavior. Specifically, dynamic layer gating
benchmark - arxiv:2611.02790 · cs.AIMultiMental-IAS: A Multimodal Driver Mental State Analysis Dataset for Context-Aware Non-IID EvaluationBaran Can Gül, Nagendra Pal Singh, Anjana Arvind Naik
Robust driver-state modelling requires synchronized physiological, behavioural, and contextual signals, yet existing open resources typically target a single cognitive state, omit continuous vehicle context, or do not define evaluation protocols that expose inter-subject heterogeneity. We present Multimental-IAS, a multimodal driver mental-state dataset collected in a cockpit-in-the-loop simulation built on CARLA. It contains vehicle telemetry with scenario-context channels, four-channel EEG and three-channel PPG from a Muse 2 headband, and cabin-facing video from 23 participants, recorded in
benchmarkevaluation protocol - arxiv:2611.02784 · cs.LGHLSClaw: Bringing Vibe Coding to High-Level Synthesis through Skill-Based Agentic HarnessJinming Ge, Shang Liu, Jiawei Liang, Shangkun Li +4
LLM agents are rapidly shifting software development toward intent-driven \emph{vibe coding}, yet their effectiveness drops sharply in high-level synthesis (HLS), where code optimization must jointly consider design structure, directive search, and downstream hardware constraints. Existing HLS optimization approaches, including both heuristic-based methods and recent LLM-assisted pipelines, typically operate at fixed abstraction levels or in weakly coupled stages, limiting their ability to leverage the strong interplay between source code restructuring and directive search outcomes. Our key in
agentllm agentagenticbenchmark - arxiv:2611.02782 · cs.CVHistory Forcing: Distilling History-Dependent Continuations for Long-Horizon Video World ModelsHangyi Kuang, Chengzhi Wang, Yefei He, Weijie Wang +7
Few-step autoregressive video diffusion enables efficient generation, yet maintaining consistency over long rollouts remains challenging, especially when subjects or scenes reappear after going out of sight. Despite progress in fidelity and rollout robustness, existing distillation methods do not fully exploit richer historical information to constrain the student's current prediction. To address this gap, we introduce \textbf{History Forcing}, a unified on-policy distillation framework that strengthens historical supervision for few-step autoregressive video generators. The framework uses the
world model - arxiv:2611.02780 · cs.CVLookahead Prediction-Guided Adaptive Visual Refocusing for Interleaved Multimodal ReasoningMengzhao Wang, Yanli Ji, Peng Ye
Interleaved multimodal chain-of-thought reasoning improves the visual reasoning capabilities of multimodal large language models by dynamically incorporating visual evidence during reasoning. However, existing methods primarily select evidence from historical context. When the reasoning focus shifts, history-guided evidence may remain focused on prior regions and fail to cover the visual information required by the next step. We refer to this lag between visual evidence and subsequent reasoning demands as Temporal Evidence Myopia. To address this issue, we propose Lookahead Prediction-guided A
benchmark - arxiv:2611.02779 · cs.LGStop Handcrafting Your Engine! GenDB-V: Towards AI-Generated and Formally Verified Database SystemsJiale Lao, Immanuel Trummer
For over 50 years, the database community has developed a wide range of successful engines to answer two fundamental questions: how can database systems run faster, and how can we ensure their correctness? Experience with these engines shows that performance depends on hardware and workloads. As both change over time, we repeatedly redesign and rebuild database systems, a time-consuming and expensive process. Correctness remains another long-standing challenge: recent studies continue to uncover wrong-answer bugs even in mature database engines, despite years of development and testing. To add
ai agent - arxiv:2611.02777 · cs.ROBeyond Policy Patches: Distilling RL Specialists into Vision-Language-Action PoliciesZhennan Jiang, Tingying Xie, Shilong Ji, Shu'ang Yu +10
Vision-language-action (VLA) policies can handle noncritical stages of long-horizon manipulation yet fail during short precision-critical phases. Lightweight reinforcement learning (RL) specialists can address these failures without directly applying RL to the full VLA, but retain auxiliary controllers and switching mechanisms at deployment. We study how to transfer their capabilities into a single flow-matching VLA. Direct imitation of specialist rollouts can reproduce action jitter and leave the VLA unprepared for states encountered during its own execution. We introduce REPAIR, a value-guid
vision-language-actionvlamanipulation - arxiv:2611.02775 · cs.CVRisk-Constrained Evaluation of Financial Document Recognition: A Formal Framework for Evidence, Critical Errors, and Selective AutomationSubodh Lonkar
Financial document recognition is commonly evaluated using character and word error rates, which do not adequately represent the consequences of errors in financial workflows. This paper formalizes financial document recognition as a risk-constrained evidence-recovery problem. It defines typed, evidence-linked financial facts and a multi-layer quality model spanning text, layout, tables, fields, semantic relations, calibration, robustness, and operational performance. It introduces non-compensatory acceptance gates based on upper confidence bounds for critical-field errors, and formulates sele
benchmark - arxiv:2611.02774 · cs.LGDIKT: A Diagnostic and Interpretable Knowledge Tracing Framework for Prediction and Model-based Learning AnalysisYi-Qiang Wua, Shaohui Chen, Hanlin Li, Baolin Yi
Knowledge tracing is a key component of intelligent educational systems. However, most existing methods mainly focus on response prediction, providing limited support for explaining predicted outcomes and improving students'learning states. This paper proposes Diagnostic and Interpretable Knowledge Tracing (DIKT), which integrates response prediction, skill-level diagnosis, and model-based learning analysis. DIKT generates bounded cognitive-state trajectories and decomposes the final response probability into skill-wise probabilities. The trajectories reveal students'evolving strengths and wea
benchmark - arxiv:2611.02769 · cs.CVUnlocking the Potential of Vision Foundation Models for Multi-Task Dense Prediction via Mixture of Task AdaptersYunzhi Zhuge, Xinzhuo Yu, Sitong Gong, Lu Zhang +2
Multi-task dense prediction seeks a single, coherent representation that supports simultaneously discriminative semantic recognition and regularized geometric/structural inference. For vision foundation models, pretrained semantic invariances conflict with the frequency and smoothness biases of dense objectives; joint fine-tuning and single-route adapters therefore compress competing gradients into one update channel, causing negative transfer and unstable cross-task agreement. We frame this as two-level adaptation: task-generic alignment that preserves priors, and task-aware specialization th
benchmark - arxiv:2611.02767 · cs.CVGaussFlaw: Fine-Grained Perceptual Diagnosis for 3D Gaussian SplattingShaocheng Shen, Zihan Zheng, Kejin Li, Lu Liu +3
3D Gaussian Splatting (3DGS) enables photorealistic novel- view synthesis with real-time rendering, yet fine-grained di- agnosis of its rendering artifacts remains challenging. These artifacts are typically spatially localized and exhibit diverse degradation patterns, making them difficult to character- ize using global quality metrics and image-level assessment benchmarks, which provide no explicit indication of where they occur or what types of artifacts they represent. To ad- dress this limitation, we introduce GaussFlaw-80K, the first large-scale dataset for fine-grained perceptual diagnos
benchmark - arxiv:2611.02746 · cs.LGReplay for LLM Continual Learning with Multi-Teacher On-Policy DistillationYiqiao Qiu
Multi-teacher on-policy distillation (MOPD) distills domain experts into one student but needs every expert at once. When domains arrive one at a time and earlier experts are gone, the previous-stage student can teach replayed prompts of earlier domains inside every stage. Some pipelines instead store every stage snapshot for a final multi-teacher recovery stage. On five domains with a 1.5B student and five public experts, this paper compares both with joint MOPD. It varies the replay budget and the replay signal. The signal is token-level distillation alone or with alignment of hidden states,
benchmark - arxiv:2611.02744 · cs.LGSubspace Representations for Synthetic-to-Real Transfer in 3-D Surface InspectionAbdulla Nasir Chowdhury, Naish Ibrahim, Gulpi Qorik Oktagalu Pratamasunu, Kazuhiro Fukui
This paper studies what subspace representations contribute to synthetic-to-real transfer in 3-D surface inspection, where models learn from synthetic defects and the only real data before deployment are scans of good parts. We use three classical constructions as instruments: the nearest-subspace classifier, one-class template subspaces and the generalized difference subspace (GDS). On 1,197 MVTec 3D-AD scans without target annotation we find that every decision rule ranks real defects at a similar window AUC, that image-level detection stays at chance, and that no rule keeps its in-domain th
benchmark - arxiv:2611.02738 · cs.MAMulti-Modal Multi-Agent System for Bridge Operation and Maintenance based on Large Language ModelXiangyun Luo, Yunkai Zhang, Yong Huang, Stephen Wu +1
Bridges are critical transportation infrastructure, and reliable operation and maintenance (O&M) is essential for traffic safety and long-term structural performance. However, bridge O&M faces key bottlenecks: no unified architecture coordinates heterogeneous resources; model inferences lack clear links to physical-mechanical laws; and experts' tacit knowledge remains difficult to formalize. This study proposes a knowledge-enhanced autonomy formalization methodology and apply it to construct the Bridge O&M Multi-Modal Multi-Agent System (BOMMA), which integrates heterogeneous diagnostic action
multi-agentagent system - arxiv:2611.02737 · cs.LGMoviEdit: Long-form Multi-Shot Video Editing Made EasySuhyeon Lee, Zhengang Li, Soo Ye Kim, Peihao Zhu +5
Long-form multi-shot video editing requires an intended edit to remain coherent across abrupt changes in scene, viewpoint, and subject appearance, while processing substantially longer visual context than conventional single-shot editors. We introduce MoviEdit, an end-to-end framework for scalable long-form multi-shot video editing that jointly addresses the challenges of data, modeling, and evaluation. To provide cross-shot supervision data at scale, we develop automatic data-generation pipelines spanning synchronized multi-view footage, real-world multi-shot videos, and generated multi-shot
benchmark - arxiv:2611.02736 · cs.LGFeature extraction for LSTM-based finger-joint angle estimation from sEMGGerda de Jager, Marelie H Davel, Ian Thomson
Continuous estimation of finger-joint angles from surface electromyography (sEMG) is a promising approach for controlling dexterous myoelectric prostheses. Deep time series models have been applied successfully in this context, typically relying on feature extraction rather than raw signals, given limited subject-specific data. Various time-domain and frequency-domain features can be used, extracted for a given window and step size. This study systematically investigates these three design choices for Long Short-Term Memory (LSTM) regression on the Ninapro DB2 dataset. Window sizes from 300 to
dexterousmemory - arxiv:2611.02725 · cs.AIAffToM: Benchmarking Recursive Affective Theory of Mind in Dynamic Multi-Party InteractionsJinpeng Hu, Yiduo Jin, Erqiang Wang, Jia Li +2
Understanding emotion in realistic social interactions requires more than recognizing facial expressions or assigning affective labels. The same event can elicit different emotions depending on what each participant knows, what they believe others know or feel, and how these nested mental states evolve over time. Existing benchmarks largely study higher-order epistemic Theory of Mind (ToM) and multimodal emotion reasoning in isolation, leaving recursive affective reasoning under asymmetric and dynamically changing information underexplored. Therefore, we introduce AffToM, a benchmark for affec
multi-agentpost-trainingbenchmark - arxiv:2611.02721 · cs.AIIssueLighter: An Issue Type-Aware LLM Agent for Issue Report Validity AssessmentJiawei Ran, Yuxia Zhang, Ziyang Zhao, Hui Liu
Invalid issue reports are prevalent in open source software (OSS) projects, imposing substantial burdens on maintainers who already face a high risk of burnout. Prior research has primarily focused on identifying invalid reports for specific issue categories (e.g., bug and wontfix), leaving the broader landscape of OSS issue types largely unexplored and limiting the generalizability of existing approaches. Moreover, existing methods provide little support for explaining or validating their predictions, reducing their practical usefulness for maintainers. With the rise of LLM-based agents, it r
agentllm agent - arxiv:2611.02710 · cs.CLMicroMouse: Masking Harmful Supervision in Long CoT via Relative PotentialXiaowei Zhu, Hengzhu Tang, Zhenyu Zhang, Shuohuan Wang +2
Even successful long CoTs may contain harmful supervision, motivating the need for methods that can automatically identify and mask such signals during SFT. Inspired by the MicroMouse maze-solving process, we view a long CoT as a search trajectory from the input problem toward the target answer. This perspective motivates maintaining a goal-relative measure, analogous to a MicroMouse's spatial coordinates, to quantify how close the current reasoning state is to the target answer. Building on this perspective, we propose MicroMouse, an automated selective-SFT framework for masking harmful super
benchmark - arxiv:2611.02706 · cs.LGHindsight Credit Assignment over Trees for Agentic RLPenghong Zhao, Yihao Chen, Maoheng Yuan, Yunjian Zhang +7
Reinforcement learning for long-horizon LLM agents faces a fundamental dilemma: successful experiences are essential for policy improvement, yet limited policy capabilities make such experiences difficult to discover. Sparse outcome rewards compound this challenge, leaving sampled trajectories with little relative feedback for learning. In this paper, we propose TriGRPO, a tree-based reinforcement learning framework that turns scarce successes into exploration and learning signals through hindsight credit assignment within and across rollout trees. TriGRPO constructs rollout trees through hind
agentllm agentagentic - arxiv:2611.02704 · cs.AIBeyond Answers, Traces, and Efficiency: Evidence for Adaptive ReasoningSejin Kim, Sundong Kim
Current AI systems are evaluated as reasoners through final-answer accuracy, reasoning traces, process supervision, and computational or human-relative efficiency. Yet these measures do not establish whether reasoning adapts when tasks or evidence change. We distinguish across-task adaptivity, where acquired abstractions transfer or recombine across novel problems, fromwithin-task adaptivity, where hypotheses and actions are revised in response to new observations, feedback, or counterexamples. Using ARC-AGI as a diagnostic case study rather than an alignment benchmark, we examine what existin
benchmark - arxiv:2611.02699 · cs.RORight Ranking, Wrong Scale: Reproducing the Robotic World Model's Dynamics Model and Testing the Uncertainty of Its Follow-upJoyjeet Singh
We reimplement the dynamics model of the Robotic World Model (RWM) and test the uncertainty estimate that its follow-up, RWM-U, uses to penalize reward in offline policy optimization. Our implementation matches the released model's gradients. On held-out data, training on the model's own rollouts beats one-step training at the preregistered long horizon for all ten episodes, but loses on short one-step windows. RWM-U's ensemble disagreement ranks rollout errors better than forecast step (Spearman rho = 0.90 at 100 steps), yet its spread substantially understates error: one-sigma coverage on th
world model - arxiv:2611.02698 · cs.LGGeoGuesser Bench: Benchmarking Vision-Language Agents on Interactive GeolocationHuanzhi Mao, Yichuan Wang, Michael Koo, Wenyuan Huang +4
Vision-language model (VLM) agents that act in the open world need two abilities: world knowledge to interpret what they see, and exploration to find what they cannot yet see. We study how agents combine the two through interactive visual geolocation. GeoGuesser Bench contains 282 tasks in an offline Street View environment where agents rotate, zoom, and move between real panoramas. Its three task types shift the balance from knowledge to exploration: agents (1) report the coordinates of their starting location, (2) name the enclosing street, city, state, or country, or (3) find, read, and cou
agenttool-usebenchmark - arxiv:2611.02695 · cs.CV3L-D: A Low-Light Level Dataset from a Sensor-Agnostic, Alignment-Preserving Acquisition FrameworkLéo Rueff, Maurgène Ghezal, Valentin Gorse, Astrid Dufaure +3
Supervised low-light image and video enhancement has achieved remarkable progress, yet current models remain fundamentally constrained by the limitations of existing datasets. Prevailing methods typically rely on uncontrolled ambient captures, camera parameter manipulation (e.g., exposure or ISO adjustments), or neutral density filters, failing to replicate authentic night-sky near-infrared spectral content and genuine low-photon noise distributions. Furthermore, capturing dynamic paired sequences without introducing warping artifacts or sensor-specific biases remains an open challenge. In thi
manipulation - arxiv:2611.02692 · cs.CLLLMs Cannot Be Neutral: Gender Sensitivity Across Candidates, Evaluators, and ModelsDanial Amin, Joni Salminen, Bernard J. Jansen
Large language models (LLMs) are increasingly used to evaluate candidates for professional roles, raising concerns about their sensitivity to demographic identities. We examine how candidate and evaluator gender assignments affect decisions by LLM-as-a-Judge (LaJ) models and decision language models (DLMs). In a hypothetical city council voting task, eight models evaluated 240 synthetic professional narratives from NVIDIA's US and French Nemotron-Personas datasets under independently varied gender assignments, producing 30,720 judgments. Explicit candidate gender labels increased Yes-vote prob
evaluator - arxiv:2611.02691 · cs.ROClosing the Dynamics Gap through Self-Directed Robotic InteractionXilun Zhang, Liam Achenbach, Luis Pabon, Matthew Foutter +1
Agentic Real2Sim2Real systems can reconstruct a scene in simulation from camera observations, but contact-rich dynamics such as friction and restitution cannot be observed without interaction. We call this the dynamics gap and present CUE, an agentic system that closes it by interacting with the environment itself. Given only a high-level instruction and no resets, the agent builds its own perception, identifies physical parameters, and targets interactions where its simulation is least accurate. On billiards, CUE reduces held-out ball-position error by 72% relative to the strongest baseline a
sim2realagentagentic - arxiv:2611.02690 · cs.LGHuman-Oriented World ModelSaihui Hou, Zengyun Mou, Chunshui Cao, Yongzhen Huang
Supporting a person over repeated interactions requires interpreting their current state in context and anticipating its subsequent evolution. We formulate the \emph{Human-Oriented World Model} (HOWM) as a conceptual framework centered on task-relevant human states and their evolution. HOWM connects current observations and individual history with environmental and social context to organize continuing state estimation, personalized dynamics prediction, and action support. Dynamic expectations and new observations jointly inform the current estimate, which in turn anchors personalized forecast
world model - arxiv:2611.02689 · cs.AIThe Customer Never Sees the Goal: Measuring and Detecting Vulnerability-Targeted Manipulation in LLM AssistantsBohan Jiang, Dawei Li, Yasin Silva, Huan Liu
Companies deploy large language model (LLM) assistants on their websites to advise customers. For commercial reasons, the operator can give an assistant a goal that the customer never sees. This paper focuses on operator-induced vulnerability-targeted manipulation (VTM), in which an assistant pursues such a goal by exploiting a vulnerability the customer discloses. While existing research mainly measures how manipulative a model is, we go further and ask whether VTM can be detected from the customer-AI conversation. To investigate this at scale, we introduce GAP (Goal-Anchored Probing). GAP fi
manipulation - arxiv:2611.02685 · cs.CVGeometry-Constrained Dynamic Hypergraph Convolutional Network with Contrastive Score Refinement for Skeleton-Based Action RecognitionMinh Ta Tran Nhat, Yuning Ye, Yuseok Ban
Skeleton-based action recognition is a pivotal task in computer vision; however, conventional Graph Convolutional Networks (GCNs) are often constrained by pairwise joint modeling, which fails to capture the intricate high-order interactions inherent in human movement. While recent hypergraph-based methods attempt to address this limitation, they typically rely on static structures derived from anatomical priors, lacking the flexibility to capture dynamic, action-specific joint correlations. To this end, we propose DyGCN, a Geometry-Constrained Dynamic Hypergraph Convolutional Network for skele
benchmark - arxiv:2611.02677 · cs.ROVisuo-tactile Diffusion Policy for Robotic Manipulation with Auxiliary Tactile Dynamics PredictionKaihong Ouyang, Zihao Su, Chengxuan Zheng, Weiyi Lai +1
Long-horizon robotic assembly requires reliable perception and action across grasping, alignment, and insertion. Vision captures global object pose and scene context, where as fingertip touch provides local contact cues that are especially useful as contact states change. We present a visuotactile diffusion policy that combines four-step histories from two camera views and two fingertip tactile images to condition action-sequence generation. A recurrent temporal representation integrates recent observations from both modalities. To encourage it to capture how contact evolves under action, an a
manipulationtactilediffusion policygrasp - arxiv:2611.02660 · cs.ROHeLM: Hierarchical Explicit Language Memory for Non-Markovian ManipulationGyeonghun Kim, Donghoon Kim, Minsoo Choi, Suyun Lee
Recent advances in Vision-Language-Action (VLA) models have enabled fine manipulation capabilities, yet performing tasks that require working memory remains a key challenge. Most existing VLA architectures operate under a Markovian assumption, mapping only the current observation to an action, which leads to failure in non-Markovian environments. Inspired by the human 'inner monologue', we propose Hierarchical Explicit Language Memory (HeLM), a hierarchical VLA framework that maintains task-relevant history in explicit language through factorized event detection and event-triggered memory upda
vision-language-actionvlamanipulationmemory - arxiv:2611.02659 · cs.LGParametric CAD Bench v3: Evaluating Agentic Parametric CAD Authoring from Text and Engineering DrawingsMei Chen, Yu Xia, Haipeng Mao
Producing an executable CAD script is not sufficient to establish that an agent has satisfied an engineering design request. Evaluation must also consider the resulting geometry, specified dimensions, and native CAD modeling history. We present Parametric CAD Bench v3, a 100-task benchmark comprising 30 text-to-CAD creation tasks, 30 text-to-CAD create-and-edit tasks, and 40 tasks that reconstruct mechanical parts from engineering drawings. Agents operate in a pinned FreeCAD environment and submit native CAD artifacts to a separate programmatic verifier. The primary metric is mean continuous t
agentagenticagent systembenchmark - arxiv:2611.02651 · cs.CVProbe-Guided Gradient Balancing for Multimodal LearningVu Vo, Haytham M. Fayek, Thuy Nguyen
Jointly trained multimodal networks frequently underutilize weaker modalities, as faster-learning modalities dominate the shared fusion objective and suppress gradients reaching slower ones. Most existing approaches throttle dominant modalities, whereas directly boosting weaker modalities often fails because monitoring signals derived from the joint loss remain dominated by stronger modalities. We propose Probe-Guided Gradient Boosting (PGGB), a gradient modulation approach that boosts weaker modalities while preserving the strength of dominant ones. PGGB attaches lightweight linear classifier
benchmark - arxiv:2611.02648 · cs.CVReinSeg: Retrieval-Guided Bounded Multimodal Refinement for 3-D Road Anomaly SegmentationHaifa Zhang, Yijing Wang, Zhiqiang Zuo
3-D road anomaly segmentation is essential for risk-aware perception in autonomous driving. Its central challenge is to rank sparsely observed hazards ahead of abundant normal road structures while preserving point-wise metric localization. LiDAR provides accurate 3-D point support, whereas surround-view imagery offers complementary semantic context for separating unexpected hazards from challenging normal structures. We propose ReinSeg, a retrieval-guided bounded multimodal refinement framework for point-wise road anomaly segmentation. Occupied LiDAR bird's-eye-view cells retrieve scene-speci
memorybenchmark - arxiv:2611.02644 · cs.MACoati: Host-Grounded Execution for Installable Mobile GUI AgentsYudou Ying, Pu Shi, Hanlei Shi, Yuhua Zheng
Mobile GUI agents can carry out tasks across apps, yet most run on a separate computer that drives a phone over the Android Debug Bridge (ADB), out of reach of ordinary users and of the apps through which most people reach Web services. We present Coati, a source-available agent that users install as an Android app. Its runtime on the phone, the host, records what a model alone cannot see: the effect of each action at its target, result screens across apps, and the progress of repeated steps. VerifyLoop checks action outcomes and completion claims against host evidence and reports a completion
memoryagentbenchmark - arxiv:2611.02640 · cs.AIInterFuse: Interaction-Guided Adaptive Fusion of Heterogeneous Models for Payment Risk RankingHaoyu Jiang, Weiyu Han, Ze Zhao, Jie Sheng +4
Payment platforms use view-specific models trained on transaction attributes, behavioral histories, and relational structures to rank transactions under limited review budgets. However, existing fusion methods often focus on prediction scores or interactions among model representations, with less attention to how these interactions should guide aggregation using each model's original prediction and input coverage. To address this limitation, we propose InterFuse, a fusion framework that uses a shared fusion state from cross-model interaction to guide transaction-adaptive aggregation. InterFuse
benchmark - arxiv:2611.02633 · cs.AIMeasuring AI Agents as Teammates with Collaborative IntelligenceSean Wu, Aneesh Pappu, Jiaru Zou, Yubin Kim +5
Progress in AI is often measured by what a single model can do on its own. Yet teams of AI agents are starting to make breakthroughs and discoveries in science, medicine, mathematics and software engineering. As AI begins to work in teams of agents and people, the ability to collaborate becomes a core part of their intelligence. Research on human teams has measured this ability for decades. Inspired by this, we introduce the Collaborative Intelligence Benchmark, which measures how well an AI agent works as a teammate. It contains 240 tasks across 8 dimensions of teamwork, from refusing an unsa
agentai agentbenchmark - arxiv:2611.02629 · cs.RODexArena: A Unified Benchmark for Long-Horizon Dexterous ManipulationHaonan Ping, Jianshu Hu, Yimin Wang, Yi Zhu +11
Long-horizon dexterous manipulation remains a fundamental challenge for robot learning, requiring the composition of multiple skills, precise bimanual coordination, generalizable visual understanding, and robust skill transitions. Yet existing manipulation benchmarks largely emphasize isolated skills or support only a limited range of end-effectors, making these capabilities difficult to evaluate systematically. We introduce DexArena, a benchmark designed to evaluate learning and generalization in long-horizon bimanual dexterous manipulation. DexArena features 10 long-horizon tasks and 30 stag
manipulationdexterousbenchmark - arxiv:2611.02626 · cs.RONeural Integral Control Barrier Functions for Agile Collision AvoidanceDasari Raj Vamsi, Vrushabh Zinage, Efstathios Bakolas, Rohan Chandra
Control barrier functions (CBFs) are a standard way to keep a planner inside a safe set, but they assume control affine dynamics, a barrier function given in closed form, and a known nominal controller. Integral control barrier functions (ICBFs) remove the first assumption and absorb input constraints, yet they still assume a given nominal controller and a barrier function. In this work, we present an approach to learn all three from data, using neural ordinary differential equations (ODEs) for the dynamics, imitation for the nominal controller, and supervised regression for the barrier functi
benchmark - arxiv:2611.02624 · cs.AIWidening Harm Evaluation Coverage with a Judge that Reads the WorldPrabhpreet Singh Setia, Bhavesh Neekhra, Debayan Gupta
An agent acting in a stateful environment can leave evidence of harm in two places: the record of what it did and said, or the world it changed. Safety evaluations for discovering harm examine one or the other in limited ways, using either an LLM judge over the trajectory or deterministic checks over the final state. Although LLM-based verifiers that inspect the environment exist, they have been used to verify task completion, not to detect and confirm harm in a safety setting. We address this gap with an LLM that has read-only access to the environment state and is given a harm taxonomy. We e
agentagenticjudge model - arxiv:2611.02613 · cs.ROMAPO: Mixture of Experts with Alternating Policy Optimization for Universal Humanoid Motion TrackingSeonghyeon Lim, Hyeonwoo Lee, Mincheol Kim, Kimin Lee +1
Universal humanoid motion tracking is essential for enabling humanoid robots to perform diverse tasks in human environments. To capture the wide diversity of human motions, mixture-of-experts (MoE) policies have been widely adopted, leveraging dynamic routing to enhance policy expressiveness. However, we identify that divergence between old and updated policies can be misestimated during policy-gradient updates of MoE policies, due to their dynamic routing. Because most policy-gradient algorithms regulate policy updates based on the divergence, this misestimation can lead to either overly cons
humanoid - arxiv:2611.02612 · cs.AIScalaris: A Scalable Enterprise Memory Architecture for Conversational AI AssistantsDevarsh Gandhi, Sarthak Pawar, Hemprasad Badgujar, Vaibhav Rathore +2
Enterprise conversational systems require memory to personalize interactions, retain information across sessions, and maintain conversational coherence. At enterprise scale, however, memory must not only retrieve relevant information accurately but also provide it efficiently to the answer-generation model. Existing memory systems face a trade-off between retrieval accuracy and context efficiency. Retrieving more context can improve recall but increases inference cost and latency. We present Scalaris, a scalable enterprise memory architecture that addresses this trade-off through a unified lon
memorymemory architecturebenchmark - arxiv:2611.02609 · cs.LGWhitened Advantage Summation Policy Optimization: Correlation-Aware Aggregation for Multi-Reward Reinforcement LearningOlaitan Olaleye, Jayson Meribe, Temidayo Olaleye
Reinforcement learning for language models often combines several rewards by normalizing each one and adding them with fixed weights. When the rewards are correlated, the variation they share is counted once for every reward that carries it. In terms of the ZCA-decorrelated components of the reward signal, the weighted sum applies weights $Σ^{1/2} w$ rather than the chosen weights $w$, where $Σ$ is the reward correlation matrix. We study WAS-PO, which whitens the normalized advantages with $Σ^{-1/2}$ before adding them, so that the chosen weights apply to those components. We derive the scalar
tool calling - arxiv:2611.02605 · cs.AIStateFence: Task-State Protection Against Evidence-Chain Contamination in Blue-Team LLM AgentsHaowen Dai, Alejandro Guerra-Manzanares
Indirect prompt injection creates a distinctive risk for blue-team Large Language Model (LLM) agents in Security Operations Centers (SOCs). Malicious instructions can be embedded in telemetry, tickets, retrieval results, or memory and then propagate through alert triage, incident summarization, response recommendation, and Indicator of Compromise (IOC) extraction as locally plausible intermediate decisions. We characterize this failure mode as evidence-chain contamination of protected task state. The main challenge is task dependent: a suspicious string may be malicious control text for summar
memoryllm agent - arxiv:2611.02603 · cs.CVBGD-SF PolySegNet: boundary-guided dynamic selective fusion network for robust polyp segmentation in colonoscopy imagesNishi Kanta Paul, Shamia Maherin, Morsheda Akter, Moinul Hossain +1
Effective polyp segmentation in colonoscopy images is crucial for early detection of colorectal cancer, a leading cause of cancer-related death worldwide. While current deep learning models perform well on traditional Dice benchmarks, border segmentation remains difficult. Most methods use boundary information only as an additional loss signal during training, which prevents the network from explicitly learning border-guided feature selection or mask refinement. This paper introduces BGD-SF PolySegNet, a boundary-guided encoder-decoder architecture that controls three phases of the segmentatio
benchmark - arxiv:2611.02601 · cs.AIParallelism-Aligned Pruning and Round-Aware Context Reuse for On-Device Multimodal AgentsPeipei Jiao, Mingkang Liu, Weiming Meng, Lichen Pang +2
On-device multimodal agents are limited by two deployment mismatches: pretrained attention structures may not divide evenly across the compute cores of an edge accelerator, causing irregular partitions; and standard prefix caching keeps historical visual tokens in the persistent KV cache, whose capacity grows with the number of rounds. We present PARC, which combines parallelism-aligned structured pruning with lifetime-aware KV-cache reorganization. At the model level, it uses first-order Taylor importance and grouped-query-attention-aware selection to make attention dimensions divisible by th
memory - arxiv:2611.02600 · cs.ROW2A: Distilling World Priors into Action Models for Real-Time Robotic ControlXiaopeng Zhang, Yongjin Mu, Yueyang Weng, Chengzhi Zhao +1
World Action Models (WAMs) have recently shown great promise in robotic control, benefiting from the future imagination capability of large-scale pretrained video models. However, most existing WAMs require explicitly imagining future states at test time, incurring substantial computational overhead and latency that make them unsuitable for real-time robotic control. In this paper, we propose W2A, a two-stage distillation framework for real-time robotic manipulation that distills intermediate denoising representations from a video diffusion model into an action model during training, while ent
manipulationlibero - arxiv:2611.02593 · cs.CVAVID: A 5T fMRI Dataset for Benchmarking Auditory-induced Visual Mental Imagery DecodingShiqi Shen, Shurui Li, Xilin Zhang, Yuanning Li
Visual mental imagery (VMI) decoding uses brain activity to reconstruct internally generated visual scenes, providing a potential pathway for externalizing internal experiences and developing neuroprosthetic communication interfaces. However, progress is limited by the lack of dedicated, training-scale VMI datasets. Current VMI paradigms often rely on visual prompts or short-term memory images, introducing visual and mnemonic confounds that make it difficult to disentangle neural activity associated with endogenous image construction from signals driven by prior or concurrent visual input. Fur
memorybenchmarkevaluation protocol - arxiv:2611.02587 · cs.AIThey Know Where It Is, But Not What It Says: A Bilingual Hallucination Benchmark of Hadith Recall in Open LLMsDayam Nadeem, Safdar Mustafa, Ameen Ur Rehman, Shahab Saquib Sohail +2
Large language models are increasingly consulted for religious knowledge, including the recall of specific Hadith (recorded sayings and actions of the Prophet Muhammad, peace be upon him). Unlike open-ended generation, Hadith recall is a high-stakes, verifiable task: a reference (collection, chapter, and Hadith number) has exactly one correct text, and any deviation from it constitutes a fabrication that can be mistaken for authentic scripture. This study introduces a bilingual (English/Arabic) benchmark that probes five open-weight, locally-deployable language models (Mistral-7B, Gemma3-4B, J
retrieval-augmentedbenchmark - arxiv:2611.02585 · cs.CLFrom Language to Country to Valence: A Multilingual Audit of Stereotype Bias Across LLMsZihao Sun
Large language models (LLMs) are increasingly considered for education, recruitment, workplace decision support, and other consequential applications, making pre-deployment social-bias detection essential. Yet major benchmarks are English-centred or translated, focus mainly on negative stereotypes, and may miss linguistic and cultural variation. We investigate gender, racial, and religious bias using 12,000 locally written Reddit posts in English, Spanish, French, and German. Only search terms are translated, preserving local context. The dataset is balanced by language and category; English g
benchmark - arxiv:2611.02583 · cs.AIBasisKV: Value Latents for Communication and Key RoutingLixun Zhang, Zishan Shao, Kangning Cui, Yu Liang +2
Long-context large language model (LLM) inference is increasingly constrained by the cost of maintaining an ever-growing KV-cache. Low-rank K/V projections provide a promising way to compress the KV-cache with minimal quality loss, but prior work has largely optimized KV-cache compression in isolation, leaving distributed inference, especially with tensor parallelism where attention computation is partitioned across multiple GPUs, largely underexplored. We make two observations: (i) tensor-parallel communication is an important cost when deploying low-rank KV-cache compression in distributed i
long-context - arxiv:2611.02581 · cs.LGSMILE: A Structured Multi-Agent Clinical Reasoning for Liver Ultrasound UnderstandingHaris Ghafoor, Minjee Seo, Seonaeng cho, Keungmo Yang +2
Automated liver ultrasound report generation improves the diagnosis of liver diseases by producing clinically structured descriptions of anatomical and disease-related findings. However, current generative vision-language models (VLMs) often rely on a single reasoning pathway that maps ultrasound appearance directly to textual descriptions. In addition to the inherent vision-language modality gap, this design is limited for liver ultrasound because disease interpretation is constrained by subtle visual evidence, heterogeneous disease patterns, and limited paired supervision. To address these l
agentmulti-agent - arxiv:2611.02576 · cs.LGBeyond $L_\infty$-Norm Sets: Synthesizing Interpretable Non-Uniform Uncertainty Sets for Evaluation-Time State-Adversarial RobustnessXulin Chen, Ruipeng Liu, Garrett E. Katz
Deep reinforcement learning (DRL) agents are vulnerable to perturbations of their observed state inputs. Prior works often constrain state adversaries with $L_p$-norm uncertainty sets and evaluate policy performance under the resulting perturbations. However, in continuous control tasks like robotics control, different state dimensions have distinct physical meanings and influence policy behavior to different degrees, which $L_p$-norm sets fail to capture or explain. Moreover, it is important to identify when and along which state dimensions a policy is vulnerable, independently of the algorit
policy evaluation - arxiv:2611.02571 · cs.LGOrganizational Learning in Materials Discovery Campaigns: Which Databases, Calculations, and Agent Teams Pay OffKinston Ackölf, Taylor D. Sparks
More databases can make a materials campaign worse. Through an organizational lens, we trace each source's contribution. One band-gap campaign ranks seven anion families from archived measured gaps, one vote per source. In the first half of its budget, all seven pooled rank worse (rank agreement 0.737; 1 = measured order) than the best pool of three (0.851). As a stability check, choosing the pool on half the runs and testing on the rest kept the gap. Splitting Materials Project afterwards points to its transition-metal oxides and fluorides. Two databases from one pipeline add nothing over one
agent - arxiv:2611.02568 · cs.CVSeMo3D: Unifying 3D Generation, Editing, and Completion via Selective ModelingXiaokun Sun, Jian Yang, Zhenyu Zhang
3D generation, editing, and completion are typically treated as regenerating the complete 3D asset under task-specific constraints, even when only a local region needs modification. This leads to task isolation, inconsistency in content that should remain unchanged, and dependence on task-specific paired data. We address these issues with a key observation: the three tasks all synthesize a designated region while keeping the rest intact---the full region for generation, the target region for editing, and the missing region for completion. We term this paradigm Selective Modeling and build SeMo
benchmark - arxiv:2611.02566 · cs.LGCR-Bench-v2: Evaluating the Real-World Utility of AI Code Review AgentsKristen Pereira, Neelabh Sinha, Rajat Ghosh, Debojyoti Dutta
The rapid growth of automated code generation has created a pressing need for automated code review. Yet code review resists full automation: it depends on context beyond the \textit{diff}, on the author's intent, and on judgments about design and maintainability on which even expert reviewers disagree. Existing code review benchmarks lack standardization and rely largely on noisy human review comments scored with text-similarity metrics, which reward surface overlap and fail to measure precision. We introduce \texttt{CR-Bench}, a defect-centric evaluation framework that recasts SWE-Bench issu
agentbenchmarkevaluation framework - arxiv:2611.02561 · cs.AIEEFA: A Framework for the Ethical Evaluation of AI AgentsSwathi Pallikala, Yashas Hariprasad
The rapid deployment of autonomous AI agents across consequential domains including healthcare, legal services, nancial management, and enterprise workows has created an urgent governance challenge. While the eld of AI evaluation has produced sophisticated benchmarks for measuring agent capability, these frameworks measure exclusively what agents can do rather than how they do it. The ethical dimensions most critical to responsible agent deployment, in- cluding respect for human autonomy, harm avoidance, transparency and non-deception, privacy compliance, and accountability, are systematically
agentai agentagent systemagent benchmarkbenchmarkevaluation framework - arxiv:2611.02555 · cs.ROSequential Recovery Updates for Learned Controllers under Successful-Trajectory PreservationZongtan Li, Tianming Liu, Ye Zhao
A competent learned controller already recovers from many disturbances. Improving it means learning a correction, and learning when not to apply it. Freezing a skill's parameters does not settle the second half. In a controller that routes among frozen skills, a new entry rule can redirect the states an old skill succeeded on. We therefore constrain entry instead of parameters. Each update searches for a feedback skill at a measured failure. It fits an entry rule that reads observations and rejects every state on an accumulated successful trajectory. The skill is admitted only if the composed
humanoid - arxiv:2611.02550 · cs.CLRe$^3$Rank: Listwise Reranking with Reasoning-Conditioned Retrieval HeadsRuiyi Yan, Zhuoyuan Mao, Yugo Murawaki, Yiwen Guo
Attention-based rerankers score each candidate by the attention mass that retrieval heads route to it from the question, but the question, serving as a fixed query, carries limited information for locating relevant evidence. We propose Re$^3$Rank, a listwise reranker built on reasoning-conditioned retrieval heads (RRHeads). A generator first writes a reasoning trace over the question and the candidates, and a scorer then uses the trace tokens as the attention query of a few selected RRHeads, reading relevance as continuous scores. We present a straight-through estimator to propagate the rankin
benchmark - arxiv:2611.02549 · physics.opticsDeterministic and dynamic synthesis of optical skyrmions through an ultrathin multimode fibreR. E. Bautista Gonzalez, R. Mouthaan, A. McIntosh, K. Dholakia
Topological light fields offer powerful opportunities for information encoding, transport, and manipulation, yet their deterministic generation in complex scattering media remains an outstanding challenge. In highly multimode fibres, strong mode coupling and polarisation mixing can obscure the precise control required to synthesise structured vectorial states, including optical skyrmions. Here, we demonstrate the deterministic generation and dynamic reconfiguration of arbitrary optical skyrmions through an ultrathin multimode fibre. By combining vectorial transmission-matrix control with a pha
manipulation - arxiv:2611.02546 · cs.ROEnergy-Based Policy OrchestrationWeirui Ye, Yen-Jen Wang, Elvis Hsieh, Zhiyang Dou +5
Hierarchical control scales robot behavior by separating task-level planning from low-level control. In dynamic, high-dimensional tasks, however, a fixed planner-to-policy interface can fail when contacts, balance, or disturbances evolve faster than high-level policy updates. Execution therefore requires choosing not only which low-level policy to run, but also which policy-specific goal should be given to it. We formulate this problem as policy orchestration: runtime selection of a policy-goal pair that is likely to remain feasible while tracking the planner target. We develop EPO, an energy-
manipulationhumanoidsim-to-real - arxiv:2611.02545 · cs.CLFindLink: Learning to Assess Scientific Connections from Source-Grounded PremisesJingxuan Wei, Song Tan, Xiaohan Liu, Chao Sang +5
Scientific discovery from the literature requires assessing connections among findings obtained under different study conditions. Source attribution makes individual findings traceable, but does not determine what they jointly support. We introduce GI-TraceBench, a benchmark of 10,282 questions in gastrointestinal oncology that unifies provenance question answering, knowledge discovery, and hypothesis evaluation through shared Paper, Claim, and Evidence Span records. We further propose FindLink, a two-stage framework for provenance-grounded scientific reasoning. Premise Grounding constructs so
benchmark - arxiv:2611.02540 · cs.LGPixelBench: Can LLMs Understand and Generate Images Pixel by Pixel?Haoquan Zhang, Feiyang Liu, Juanxi Tian, Wanhua Li +2
Large language models (LLMs) can generate visual assets through text-based formats such as SVG and CAD, but these symbolic representations describe images using high-level primitives rather than explicit raster values. We ask whether LLMs can instead generate pixel images directly through their existing vocabulary, without a learned image tokenizer or decoder. We introduce PixelBench, a benchmark for text-to-pixel (T2P) and image-to-pixel (I2P) generation across multiple resolutions, together with a complementary pixel-understanding evaluation. Models generate an RGB color palette followed by
benchmark - arxiv:2611.02539 · cs.AIScientific Liquidity Agents: An Agent-Based Model of Knowledge Arbitrage in Scientific CommunitiesKoutian Wu, You Liu, Yue Liu, Zhiping Li +1
Scientific knowledge gains new uses when ideas, methods, and results move between research contexts. Understanding this process requires a model of both specialized knowledge production and the intermediary activities that connect it. We introduce Scientific Liquidity Agents (SLA), an agent-based model of knowledge arbitrage in scientific communities. SLA represents domain specialists and epistemic arbitrageurs within a shared environment of research problems, researcher interactions, and scientific artifacts. Individual decisions connect to collective state through a belief-desire-intention a
agentic - arxiv:2611.02536 · cs.AIFORGE: Fault-Oriented Rework Code Generation for Executable Interlocking AssemblyFuqiang Wang, Jiahao Fu, Ren Hefeng, Yining Qian +4
Large language models have demonstrated strong capabilities in structured reasoning and code generation, yet identifying an assembly fault does not guarantee the ability to generate executable rework programs. In interlocking assemblies, a locally plausible repair program may violate blocking dependencies, omit required recovery operations, or terminate in an incorrect configuration. We identify this challenge as fault-conditioned executable rework generation, where models must jointly reason about structural failures and generate programs that restore constrained assembly states. To benchmark
benchmark - arxiv:2611.02526 · cs.CVINTERACT: Learning Interaction-Aware World Dynamics from Driving Videos for Autonomous DrivingZhan Chen, Songyan Zhang, Wenhui Huang, Zirui Li +4
Safe autonomous driving requires more than recognizing the current scene: the ego vehicle must anticipate how surrounding traffic participants will evolve under its own motion. However, existing end-to-end driving models typically encode scenes into dense visual or BEV features, leaving object-level interactions implicit. We propose (INTERACT INTERaction-aware Action-Conditioned Trajectory learning), a three-stage framework that learns a slot-based object-centric world model from unlabeled driving videos and transfers it to trajectory planning. INTERACT first learns recurrent slots that group
world modelaction-conditioned - arxiv:2611.02525 · cs.ROTouching Projected Agents through the DAT Screen -System Design and Questions of Interaction Boundaries-Takaya Tamamoto, Tomoko Yonezawa
Touching a virtual agent involves more than detecting a finger on a display: the contact surface can deform and resist the user's push, and the agent can respond to that action. We describe the Desktop Amid Touch (DAT) screen, which combines spatial projection onto a semi-transparent, deformable mesh with touch sensing and motorized control of resistance to indentation. An initial prototype detected a single touch and estimated its position and depth from tension changes around a flexible projection screen. A subsequent prototype added motorized tension control to vary the screen's resistance
agent - arxiv:2611.02521 · cs.LGOn the Convergence of Alternating Regret Matching AlgorithmsAnh Duc Nguyen, Ioannis Anagnostides, Brian Hu Zhang, Tuomas Sandholm
Alternating updates, that is, where each player is updated in turn, tend to typically perform better in regret matching algorithms than simultaneous updates. Alternation within the regret matching family has long been a central ingredient of practical solvers for zero-sum games. Yet our theoretical understanding is lagging behind. In this paper, we address this gap by developing a systematic way of analyzing such algorithms. Our framework centers on an alternating analog of the RVU property that we introduce, which we use to obtain the optimal $T^{-1}$ average-iterate convergence within this f
benchmark - arxiv:2611.02519 · cs.CVMedHEB: Benchmarking Medical Embeddings Across Heterogeneous Clinical EvidenceYingshu Li, Shaoyang Zhou, Zhanyu Wang, Yunyi Liu +5
Clinical evidence is inherently relational: medical images are interpreted with prior studies, cross-modal examinations, reports, and localized findings, and retrieving relevant evidence is therefore central to clinical interpretation. Recent foundation models increasingly encode these heterogeneous sources into shared embedding spaces, yet existing embedding evaluation often focuses on isolated matches, leaving unclear whether current models retrieve relevant evidence across changes in modality, appearance, and granularity. We introduce MedHEB, a unified benchmark with 107 retrieval tasks fro
benchmark - arxiv:2611.02517 · cs.ROA Study on the Effects of Finger Movements on a Humanoid Hugging RobotDaigo Suhara, Tomoko Yonezawa
This study aimed to examine the impact of finger movements in hugging robots on users by experimentally investigating how the presence and patterns of such movements affect users psychologically. The results suggest that the presence and patterns of finger movements influence the sensations users seek from the robot; specifically, repetitive movements and asynchronous movements across multiple fingers tended to enhance these sensations. Additionally, simultaneous movement of all fingers and gentle hugging tended to evoke a greater sense of care. Conversely, no differences were observed regardi
humanoid - arxiv:2611.02516 · cs.AICumulative and Audited Guardrail Tool Evolution for LLM Agent SafetyXiao Yang, Aishan Liu, Siyuan Liang, Zonghao Ying +4
Large language model (LLM) agents increasingly operate in open-world environments where safety risks cannot be fully predefined. However, most existing guardrail mechanisms remain static, relying on fixed rules, prompts, or classifiers that generalize poorly to emerging risks. Here we present EVOLVE, a tool-level self-evolving defense framework that formulates agent security as a process of cumulative security evolution. Inspired by how human societies transform experiences into transferable protective knowledge, EVOLVE externalizes security knowledge into executable guardrail tools that are c
embodiedagentllm agentautonomous agentagent systemself-evolving - arxiv:2611.02511 · cs.LGWhen Prediction Shapes the Future: A BlueSky Agenda for Reflexive Time Series AnalysisTao Xie, Yiqun Zhang, Yiu-ming Cheung
Time-series models are increasingly deployed in operational systems (e.g., inventory management and clinical care) where predictions trigger interventions that alter future observations, producing records we term AI-shaped trajectories. This sequence-prediction coupling deepens as such systems evolve into agents that autonomously predict, act, and remember across long horizons. Such AI-shaped trajectories support immediate performance monitoring, but turn ambiguous once reused as passive data. An algorithm may then misread a prediction-prevented stockout as naturally stable demand, or an avert
memoryagent memoryagentautonomous agent - arxiv:2611.02510 · cs.RONeuro Inspired Robotics: What Robot Learning Can Borrow from the Sensorimotor BrainRuihan Zhang, Davy Deng
Robot policies still fail at everyday manipulation: an unfamiliar task, an object never held before, a scene that changes mid-motion, a skill that should improve with use. They fail despite pretraining on over a million trajectories, pooled from dozens of robots at great expense, while a child acquires all of these abilities within a few years of learning from a single body. This survey argues that the gap is one of design, and proposes designs for the next generation of robot models, drawn from how the nervous system solves the same problems. From the body we take compliance: the policy comma
manipulationself-improvement - arxiv:2611.02508 · cs.LGAttentionIndex: Accelerating Learned Sparse Attention with a Causal Inverted-File IndexJeongmin Brian Park, Wen-mei Hwu
Learned sparse attention, used in production models such as DeepSeek-V4 and GLM-5, lets each query attend to only k keys chosen by a lightweight indexer, but the indexer still scores every key in the prefix. At one million tokens, this selection takes more than half of prefill time. Two-stage selection over contiguous blocks, as in HiSA, reduces this cost, but its recall declines as context grows at a fixed candidate budget. We present AttentionIndex, which routes each query over an inverted-file index that groups indexer keys by similarity and refines the candidates with the native indexer sc
long-contextbenchmark - arxiv:2611.02506 · cs.CVCellPlex: Benchmarking Single-Cell Spatial Proteomics Prediction from H&EZijun Gao, Yijiang Chen, Xiangde Luo, Pheng-Ann Heng
Spatial proteomics measures tens of proteins in every cell of a tissue section, but its cost confines it to small research cohorts, whereas hematoxylin and eosin (H\&E) slides exist for nearly every patient. Inferring single-cell spatial proteomics from H\&E would bring its analyses to routine pathology, and a growing number of virtual staining models pursue this goal. Each is evaluated on its own cohort, panel and pixel-level metric, and released models transfer poorly to new cohorts, leaving open which approach to build on and which proteins H\&E can reveal. We introduce CellPlex, a benchmar
benchmark - arxiv:2611.02503 · cs.AITowards Multi-Value Alignment via Value Localization in LLMsXueqi Ma, Yanbei Jiang, Xingjun Ma, James Bailey +1
Aligning large language models (LLMs) with multiple, potentially conflicting human values remains challenging, as existing multi-objective alignment methods often require costly fine-tuning and provide limited insight into the internal mechanisms underlying value-related behaviors. In this work, we investigate multi-value alignment from a mechanistic interpretability perspective by identifying value-sensitive attention heads (value heads) in LLMs. Using a probing-based value localization framework, we discover sparse sets of attention heads associated with diverse dimensions, spanning both hum
benchmark - arxiv:2611.02501 · cs.LGSpecPilot: Autonomous Algorithm Discovery and Systems Optimization for Speculative DecodingKunjun Li, Wenhao Chai, Yihao Liang, Yuezhou Hu +6
Speculative decoding accelerates memory-bound LLM decoding with a lightweight drafter, and parallel drafters extend this speedup by drafting a whole block in one forward pass. However, it remains unclear whether their released systems (stock) attain the full speedup that parallel drafting permits. We identify three challenges. First, with one request, each step accepts few tokens but reads every target weight. Second, with batching, per-request verification work roughly halves the speedup over autoregressive (AR) decoding. Third, speculative decoding is deployment-specific, and no universal op
memoryautonomous agentagenticbenchmark - arxiv:2611.02497 · cs.RORoboColosseum: Evaluating Physical Intelligence at Scale In-The-WildTan Sang Nguyen, Yi Ru Wang, Shailesh Shailesh, Gang Yang +23
General-purpose robot policies must work across robots, tasks, and environments, yet the variation that makes physical trials informative also makes them difficult to reproduce. We introduce RoboColosseum, a distributed evaluation framework that combines unified policy serving and standardized protocols across a network of locally operated robots in diverse environments. Our initial study evaluates five vision-language-action (VLA) policies and three frontier AI agents across four embodiments (SO-101, Franka, Bimanual-YAM, and G1 Humanoid). The Open Track compares models without task-specific
vision-language-actionvlahumanoidpi0gr00tfranka - arxiv:2611.02491 · cs.CLSame Sentence, Different Construct: Identity-Swapped Rationales Reveal Measurement Non-Invariance in LLM Hate Speech ClassifiersWenchang Wang, Yue Song
Large language models are increasingly used as measurement instruments for socially consequential constructs such as hate speech, and are typically validated by their aggregate agreement with human labels. A valid instrument must also apply the same construct to every group it measures. We test this requirement using 421 \hc{} templates that are instantiated identically for seven protected groups, so that only the identity term varies. Four LLM classifiers each exceed 0.95 balanced accuracy, yet within identical templates two of them miss hateful statements about women at markedly higher rates
benchmark - arxiv:2611.02488 · cs.LGEnd-to-End Stability of Online Active Learning in Large-Scale Neural SystemsShuiwen Shen, Jinhua Zhang, Chenlei Hou, Peng Shi +5
Online Active Learning (OAL) aims to extend deep learning from static inference after offline training to continuous adaptation during real-time interaction with a changing physical environment. This transition is essential for safety-critical applications, but it raises a central theoretical question: whether the online parameter update process remains stable when scaled to multilayer neural architectures. This paper develops a stability-focused OAL framework. The framework uses prediction error as a feedback signal to update a structured deep model, thereby linking online learning with Lyapu
online learning - arxiv:2611.02484 · cs.AIBenchmark of THReD: Evaluating an LLM Agent and Its Preprocessing Tools for Turning Electronic Health Records into AI-Ready DatasetsJavad Mohammad Alizadeh, Huanmei Wu
LLM agents are increasingly built into data-analysis software, yet little evidence shows how reliably they preprocess clinical data or whether the tools they invoke give correct, reproducible output. We report a benchmark of THReD, a web-based system that turns electronic health record (EHR) extracts into AI-ready datasets through a chat agent whose proposed steps run only after user confirmation. The agent benchmark has 141 tasks (100 templates) in seven categories, including goal-level requests, failure recovery, requests calling for no action, schema variation and prompt injection through t
agentllm agentagent benchmarkbenchmark - arxiv:2611.02480 · cs.AIChain on Demand: Deferred Evidence Organization for Long-Term Conversational MemoryDi Lu, Zechen Li, Hao Xue
Long-term conversational memory faces a timing mismatch: write-time summaries and graphs organize history before a future query is known, while independent retrieval can allocate a limited evidence budget to redundant or fragmented turns. We introduce Chain on Demand (CoD), a query-time evidence organization framework over an unchanged turn-level memory. Given a broad candidate pool, CoD constructs a temporary evidence chain by allocating a fixed anchor budget according to query relevance, lexical diversity relative to selected anchors, and newly covered query terms, followed by deterministic
memory - arxiv:2611.02478 · cs.LGMetaPOCS: Meta-Learned Multi-Exit Unfolding for Budget-Aware Seismic Data InterpolationBaisong Jiang, Chunxia Zhang, Xiaoli Wei, Anxiang Di +2
Seismic data interpolation must reconcile reconstruction fidelity with the computational budgets of acquisition quality control and subsequent imaging. Fixed-depth unfolding networks are optimized for a single operating point, whereas their intermediate estimates may be insufficiently trained for early deployment. We present MetaPOCS, a meta-learned multi-exit unfolding framework for budget-aware seismic interpolation. Building on a spatial--spectral POCS reconstruction backbone, MetaPOCS supervises intermediate stages and preserves the acquired traces at every exit. A Meta-GF-based optimizati
evaluation framework - arxiv:2611.02474 · cs.CVGnomon4D: Learning Where and When to Focus for Geometric Motion-Grounded World-Action ModelingLingwei Dang, Yan Zhou, Juntong Li, Zhenhao Zhang +9
World-action models (WAMs) use video-generation priors to jointly predict future observations and robot actions. Yet RGB futures entangle physical state changes with appearance, while frame-wise depth and point maps lack temporal correspondence. Existing 4D approaches either keep motion implicit in visual features or translate predicted traces into actions through a separate interface. We present Gnomon4D, a WAM that jointly generates future video, robot actions, and GeoDelta, a video-aligned 4D geometric motion field encoding the cross-frame image-plane displacement and depth change of embodi
robotwingrasp - arxiv:2611.02472 · cs.AIMAGAR: Multimodal Agents for GUI-Based 3D Mechanical Assembly and Minimal ReworkJingxuan Wei, Jiahao Fu, Fuqiang Wang, Xiaoming Wu +4
A GUI command can be correct at the interface level yet still produce an incorrect assembly. An agent must identify the relevant components, respect their structural relationships, and determine which parts should move during recovery. We introduce MAGAR-Bench, a benchmark of 4,000 multimodal samples from 26 assembly families that separates Assembly Decision, Atomic Action, and Minimal Rework. For rework, the objective counts distinct moved parts within the validated recovery candidate set for each state, rather than the number of actions. We also introduce MAGAR, a selective correction framew
agentbenchmark - arxiv:2611.02462 · cs.ROLearning Purposeful Co-designs of Bio-inspired Quadrupedal RobotsFabio Boekel, Jiatao Ding, Cosimo Della Santina
Treating morphology and control as separate design problems has constrained the efficiency and adaptability of quadrupedal robots, while current co-design approaches are often constrained by restricted design spaces. We overcome this limit by introducing Dogger, a bio-inspired quadruped with modular three-segment legs, four-bar linkages and adjustable passive knee compliance. A two-stage framework combining reinforcement learning and Bayesian optimization is developed: the morphology-conditioned control policy is trained across designs and then held fixed during task-specific design selection.
quadruped - arxiv:2611.02453 · cs.AIRecGroupMemory: Feedback-Aware Cohort Memory for E-Commerce RecommendationWendong Zhang, Weipeng Huang, Zhixin Ma, Bin Zhang +1
Cohort memory can support LLM-based recommendation planning through group-level user profiles and advice. As outcome feedback becomes available, this guidance may need to be updated while preserving the underlying observations. We present RecGroupMemory, a workflow that provides this memory to an orchestration LLM and separates memory construction from feedback-based revision. It uses feedback from the previous day to revise decision guidance, while validation checks keep selected source values unchanged. Online comparisons were conducted among Taobao users with low purchasing power selected b
memory - arxiv:2611.02449 · cs.LGTurbo-GMM: FlashAttention-Inspired GPU-Accelerated Gaussian Mixture ModelMomin Abbas, Sadia Asif, Asad Ul Haq, Kaoutar El Maghraoui
Expectation-maximization (EM) for Gaussian mixture models (GMMs) requires scoring $N$ data points against $K$ mixture components. Prior fused GPU implementations, notably Flash-GMM, eliminate materialization of the $N \times K$ responsibility matrix but traverse the component dimension twice to compute normalization terms and accumulate sufficient statistics. We introduce Turbo-GMM, a single-sweep fused EM kernel that completes responsibility normalization and sufficient-statistic accumulation within one traversal of the component dimension. By retaining the complete component axis on chip, Tu
benchmark - arxiv:2611.02440 · cs.AIHarvesting SWE Benchmarks from Wild Vibe Coding SessionsAo Lu, Yuanhao Zeng, Shaocheng Lan, Sijie Xu +4
Development teams need reproducible tasks from their own repositories and workflows to evaluate coding agents. Conversational developer-agent collaboration ("vibe coding") produces abundant coding sessions and repository checkpoints, but requirements evolve across turns, unrelated conversations co-occur with commits, and engineering efforts can span multiple sessions. We present SessionFarmer, an automated framework that turns these records into self-contained, executable software engineering tasks. It analyzes dialogue and commit history to filter unrelated sessions and group multi-turn, pote
benchmark - arxiv:2611.02438 · cs.CLAs Expected? Efficient LLM Planning via Expectation-Observation Consistency VerificationQingfu Zhu, Huanyang Zheng, Ting Liu, Wanxiang Che
LLM planning often relies on reactive executors (e.g., ReAct) that generate a single action at each step to remain grounded in the latest observations, incurring substantial inference cost. Batched executors reduce this cost by predicting multiple actions in a single model invocation, yet they are prone to compounding errors, as intermediate observations are ignored. We propose an expectation-observation consistency verification (EOCV) mechanism for batched executors, preserving grounding while improving efficiency. The mechanism generates expectations of action outcomes and compares them with
benchmark - arxiv:2611.02433 · cs.CVBlueprint2CAD: Benchmarking Foundation Models on Engineering Drawing-to-CAD Render CorrespondenceAlexander Valverde, Danny Xie-Li, Zhichao Wang, Marcos V. Conde
Engineering drawings specify mechanical parts through orthographic projections. Recognizing which part a drawing depicts underlies part search, design reuse, and digitization, yet no public benchmark measures whether vision models can do it. We introduce Blueprint2CAD, a benchmark for drawing-to-render correspondence: given an orthographic line drawing and a shaded render, usually from another viewpoint, decide whether they show the same part. Its data layer, BlueCAD-30K, applies an open drafting pipeline to 29,925 Zero-to-CAD models, adding view-aligned renders, clean and dimensioned drawings
benchmark - arxiv:2611.02432 · cs.LGDiagnosing Model--Item Interactions in Language Model Benchmarks with LSIRMJinwen Luo, Han Jiang, Ziang Xiao, Susu Zhang
Language models with similar benchmark accuracy can succeed on different items. Overall scores do not show where their performance differs or which item properties might explain those differences. We demonstrate how the latent space item response model (LSIRM) can help evaluators investigate these patterns and their consequences for model comparison. We analyze 1,348 language models on 5,761 BBH items and compare ways of presenting and scoring answers across BBH, GPQA, MMLU-Pro, and MuSR. We examine LSIRM maps, which summarize differences in performance across models and items, alongside item
benchmarkevaluator - arxiv:2611.02430 · cs.CLInteractionBench: Measuring Timing, Task Accuracy, Whisper and Volume, and Grounding in Real-Time Voice SystemsRichard Yucheng He, Chen Xu, Yihang Liu, Tairan Chen +1
Most benchmarks for real-time voice systems focus on one aspect, such as turn-taking, instruction following, paralinguistic control, or conversational memory, each with its own data and set of systems. We present InteractionBench, a benchmark and evaluation harness that measures four dimensions on the same systems from the same recorded user audio: conversational timing, spoken-task accuracy, whisper and volume, and groundedness in conversations of more than twenty turns. The benchmark contains 221 items per system in thirteen scenarios. Before any system runs, we fix the expected floor-contro
benchmark - arxiv:2611.02427 · cs.LGArchitecture-Aware Mixed Precision for Replica-Exchange Training of Geometry-Aware Neural Operators with Uncertainty QuantificationUjin Jeon, Michael Penwarden, Wenqian Chen, Guang Lin
Geometry-aware neural operators face substantial computational and memory demands when learning across complex physical domains. We introduce Architecture-Aware Mixed Precision (AA-MP), a replica-exchange training strategy that allocates numerical precision according to replica roles and architecture-specific operations. By retaining FP32 for refinement while selectively employing BF16 for exploration, AA-MP preserves predictive accuracy while improving training efficiency. Across fluid dynamics and three-dimensional structural mechanics benchmarks, AA-MP achieves comparable accuracy to full-p
memorybenchmark - arxiv:2611.02419 · cs.CVTERA: Task-Evidence Restoration for Underwater Object Detection under Bandwidth-Constrained TransmissionHeejin Kim, Joon-Seok Song, Eunil Park
Underwater object detectors are typically evaluated on pre-transmission images, overlooking additional degradation from bandwidth-constrained downscaling and compression. We introduce AquaTX, a controlled benchmark built from real underwater detection datasets with seen and held-out codecs. We also propose TERA, a metadata-blind receiver-side framework combining task-guided pixel restoration and dual-state feature correction while keeping the detector frozen. Under seen compound degradation, TERA improves mean average precision (mAP50:95, averaged over IoU thresholds from 0.50 to 0.95) from 37
benchmark - arxiv:2611.02415 · cs.LGBehavior-Discriminative Reward Shaping for Reward-Robust Reinforcement LearningZixuan Liu, Fangzheng Wu, Brian Summa, Zizhan Zheng
Reward-robust RL typically models misspecification by specifying an uncertainty set that constrains the discrepancy of plausible rewards from a base reward. However, such reward space discrepancies can be behaviorally irrelevant: under potential-based reward shaping (PBRS), many distinct rewards preserve policy ordering. This can introduce substantial redundancy into standard uncertainty sets and degrade optimization performance. We propose Shaping-Aware Reward-Robust RL, which constructs uncertainty sets over PBRS equivalence classes by projecting each reward to a canonical representative, en
benchmark - arxiv:2611.02414 · cs.AISGR-OPSD: Learning from Subgoal Plans with Graph-Regularized Credit Assignment for Agentic Reinforcement LearningShu Liu, Yang Chong, WenShi Chen, Man Lan
OPSD provides dense supervision for agentic reinforcement learning through privileged information feedback. However, existing privileged information often lacks explicit task-progression planning, while teacher-student log-probability gaps reflect shifts in model preference rather than true task progress. To address these limitations, we propose Subgoal-Guided Graph-Regularized OPSD (SGR-OPSD), a framework that combines subgoal supervision with interaction-derived evidence for turn-level credit assignment. Specifically, SGR-OPSD uses LLM-generated subgoal plans to provide dense action-level gu
agentic - arxiv:2611.02413 · cs.LGA Scientific Claim Stability Framework for Evaluating MRI Morphometry PipelinesDuy-Cat Can, Dang-Duat Tran, Anh-Khang Nguyen, Quoc-Bao Tran +5
MRI morphometry pipelines are often evaluated by measurement agreement or downstream accuracy, yet their scientific value depends on the conclusions they support. Two pipelines can produce similar feature-level results yet support different conclusions about aging, cognitive decline, or neurodegeneration. We introduce Scientific Claim Stability (SCS), a claim-centric evaluation framework that converts region-of-interest (ROI) morphometry tables into auditable claim cards and evaluates their support under changes in dataset, processing pipeline, atlas, feature family, and target. Across 10,370
evaluation framework - arxiv:2611.02412 · cs.LGA Novel Federated Learning Framework for Defending Against Gradient Inversion AttacksPengxin Guo, Ziying Huang, Shuang Zeng, Xiaodan Zhang +3
Federated Learning (FL) offers a paradigm for collaborative machine learning model training among clients, preserving data privacy by avoiding direct data sharing. However, the gradient information exchanged in FL has been shown to be vulnerable to Gradient Inversion Attacks (GIA), which can reconstruct sensitive user data. To counter these threats, various privacy-enhancing technologies like Secure Multi-party Computing (SMC), Homomorphic Encryption (HE), and Differential Privacy (DP) have been integrated into FL. While effective, these methods often introduce significant trade-offs between p
memory - arxiv:2611.02407 · cs.AINetworked Synthetic Social Engineering on the WebBryan C. Boots
Several genuine AI services may repeat one unsupported source, and correcting that source may leave later reports unchanged. We examine these networked consequences of synthetic social engineering (SSE), deliberately misleading influence intended to redirect AI recipients from authorized objectives. Our framework connects social influence, information aggregation, and misinformation correction research to observable evidence and permission boundaries. It separates originating attempts from legitimate intermediaries' work and distinguishes identity, evidentiary support, and authorization. We te
agent - arxiv:2611.02406 · eess.SYBeyond Binary Controllability: Reachability Margins for Power Systems via Scalable ScreeningAbdallah Alalem Albustami, Ahmad F. Taha, Mohamad H. Kazma
This paper studies unreachability screening for power systems, focusing on the inadequate control authority that arises under stressed grid conditions. Its main result is an infeasibility \textit{margin} for linearized differential algebraic equation models that quantifies how far a prescribed terminal target lies from the reachable set in finite time. The paper specifically develops a conic benchmark method together with two first order methods based on projected subgradients and bundle updates for scalably repeated screening. Studies on IEEE benchmark networks show reduced terminal reachabil
benchmark - arxiv:2611.02405 · cs.CVMirrorComm: A Communication-Efficient Cooperative Perception Framework via Mirrored World ModelsShenghan Luo, Yin Xu, Cixiao Zhang, Yang Wang +3
Cooperative perception (CP) requires substantial communication bandwidth to exchange sensing information among connected agents. A variety of compression techniques have been proposed to reduce transmission costs while retaining task-relevant information. However, most existing methods follow a frame-wise compression paradigm, overlooking temporal predictability and repeatedly transmitting information that could be inferred from historical context. To address these limitations, we propose MirrorComm, the first world model-based communication-efficient CP framework. Specifically, mirrored world
world model - arxiv:2611.02391 · cs.ROLate Autumn: LLM Agents Specialize a General Multi-Robot Planner for One WarehouseYimin Tang, Han Zhang, Shao-Hung Chan, Erdem Bıyık +2
No algorithm is best on all problems, but a real warehouse is a narrow problem: its layout and processes rarely change. Operators still run one general multi-robot planner at many sites, because experts cannot tune it for each one. In this work, which we call Late Autumn, we explore whether LLM coding agents running autonomous research can specialize a general planner for one site, trading generality for throughput, much as post-training adapts a pretrained LLM to one task. We study six campaigns on a lifelong planner that combines task assignment with multi-agent path finding, on a floor mode
llm agentmulti-agentpost-training - arxiv:2611.02387 · cs.LGEviCast: Forecasting Retrieval Gains Through Shared Response CalibrationYibo Wang, Jisheng Dang, Bimei Wang, Jiongpei Hou +8
Retrieval-augmented generation (RAG) extends large language models with external evidence, yet relevant evidence does not ensure that a generator can use it. Indiscriminate retrieval can degrade task performance while adding retrieval and long-context inference costs. Existing methods neither determine before large-scale execution whether a generator, task distribution, and retriever configuration is worth deploying nor diagnose multiple candidate retrievers under a shared budget. We introduce EviCast, a framework that forecasts retrieval value through interventional evidence-state decompositi
long-contextretrieval-augmentedrag - arxiv:2611.02368 · cs.AIBeyond Accuracy: How Quantization Affects Coding AgentsAreej Mehboob, Muhammad Ali Shafique, Muhammad Usman Qadeer, Hamza Farooq
Agents are increasingly used for coding tasks, but the inference cost of large language models (LLMs) limits where they can run. Quantization reduces model size, but its effects on coding agents that execute and revise programs remain underexplored. We compare BF16 with 4-bit AWQ and GPTQ for Qwen2.5-Coder-Instruct (7B, 14B, and 32B) and Devstral-Small-2507 (24B) on OJBench coding problems across three seeds. We detect no statistically significant accuracy loss for any of the coding agents. However, final accuracy can conceal substantial behavioral changes: AWQ reduces Qwen-7B one-shot solves
tool use - arxiv:2611.02366 · cs.LGDo Visual RL Gains Scale at Test Time? A Quantile View of Post-TrainingAustin Wang, Lexing Ying
Reinforcement learning (RL) post-training improves the single-sample quality of visual generative models, but deployment often provides additional inference compute through Best-of-N sampling. Whether RL gains transfer under such test-time scaling, particularly on difficult held-out tasks, remains poorly understood in the visual domain. Our central contribution is to systematically identify and characterize a failure mode of scaling transfer in visual RL: across multiple RL methods, model backbones, and evaluations spanning compositional generation, text rendering, and preference alignment, hi
post-training - arxiv:2611.02363 · cs.CVGEAR: Generative Editing for Scene-Conditioned Outfit Retrieval for VR TravelGia-Huy Dinh, Thinh-Phuc Nguyen, Thanh-Hai Nguyen, Lam-Huy Nguyen-Ngoc +4
In virtual environments, choosing appropriate outfits requires contextual harmony with surrounding scenes, yet bridging the visual gap between 3D ambient views and catalog garment assets remains challenging. This paper investigates scene-conditioned outfit retrieval for interactive 3D environments, where a user's egocentric view serves as the retrieval signal. We present GEAR (Generative Editing for Asset Retrieval), a framework that bridges the cross-domain gap by synthesizing intermediate visual queries rather than predicting direct compatibility. GEAR uses a vision-language model to describ
benchmark - arxiv:2611.02361 · cs.LGAgent Memory as Rate-Limited Filtering: Explicit State Outlasts Long Context Beyond a Reader-Dependent HorizonTianhao Guo, Chongyao Guo
Can a language model that reads the full history of an access log say whether a service may act now? On an executable benchmark of grants, withdrawals, reinstatements, expiries, groups and change windows under a known policy, scored on pairs of histories that differ only in the order or one field of the deciding events, open models of 7 to 32 billion parameters fall from 72 to 86 percent after eight or sixteen events to 48 to 55 percent after 256, while reasoning-trained 32B models reading the current state reach 96 and 98 percent. A reasoning model of 120 billion parameters trails the state b
memorylong contextagent memoryagentbenchmark - arxiv:2611.02353 · cs.CLVisCodeBench: a new benchmark dataset for Text-to-Vis reflecting real-world practiceHy Nguyen, Thomas James Elton, Ella Wilson, Andrew Reeson +3
Generating visualisations from natural language requests (text-to-vis) enables users to explore data more efficiently. While LLMs can complete simple visualisation requests, their performance on challenging ones is not well covered by existing benchmarks. We introduce VisCodeBench, an end-to-end text-to-vis benchmark with 2,672 samples from three different domains: educational, scientific, and general audience scenarios with varying levels of complexity. We benchmark three visualisation systems---Codex, Claude Code, and LIDA---paired with a range of open and commercial LLMs across seven metric
benchmarkleaderboard - arxiv:2611.02341 · cs.CVAgiWorld: An Agentic Video World Model for Building Persistent WorldsShaoheng Fang, Yiqun Mei, Hanwen Jiang, Xin Sun +6
Recent interactive video world models generate visually compelling videos from a single image, following user-provided camera motion and text prompts. However, such low-level control demands substantial user effort to realize a high-level intent, and the generated world struggles to remain persistent as the exploration proceeds. In this work, we present AgiWorld, an agentic video world model that explores and builds a persistent world from a single image and a free-form instruction. For high-level planning, our agent AgiWorld-Agent designs the world and the journey through it in a world state
world modelmemoryagentagenticbenchmark - arxiv:2611.02340 · cs.LGOn Stochastic Continuous-Time Linear System Identification for Adaptive StabilizationReza Sadeghi Hafshejani, Mohamad Kazem Shirani Faradonbeh
Learning to stabilize an unknown continuous-time stochastic linear system is a central problem in control theory and reinforcement learning. This paper numerically studies data-driven identification of such systems with possibly unstable dynamics for the purpose of stabilization. We present a method that learns a stabilizing gain from observed trajectories, and examine how the failure probability, namely the probability that the learned gain does not stabilize the system, depends on the amount of data, dimension, noise intensity, excitation, and controllability. Numerical experiments on an uns
benchmark - arxiv:2611.02338 · cs.LGMemShift: Separating Detection from Repair in LLM-Derived Personal MemoryZichuan Xu, Haochen Zhang, Yuyang Dong, Makoto Onizuka +1
AI assistants store values that a language model derived from a person's records under the person's own definitions, such as their important contacts. Keeping such a value correct takes two steps: detection, noticing that an input changed, and repair, producing the correct new value. Benchmarks that score only the final answer cannot tell which step failed, and they state every change as new text. Seeing this limitation, we present MemShift, a benchmark of five shifts: source revision, deletion, redefinition, writer switch, and export. The first three fork a synthetic world into branches that
memorybenchmark - arxiv:2611.02334 · cs.CLDoes a Norwegian further-trained model give more outdated legal values than its base model? A preregistered paired comparison of Borealis 2 and Gemma 4Eirik Botten Nicolaysen
We compare how often two released language models give outdated numeric values from Norwegian statutes and regulations: NbAiLab/borealis2-26b-a4b-preview (Borealis 2), which the National Library of Norway further trained from google/gemma-4-26B-A4B-it (Gemma 4), and Gemma 4 itself. The study was preregistered (doi:10.5281/zenodo.23858965) and uses the closed norregel benchmark, scored deterministically against a value history. On the 411 of 503 primary items where both models gave an interpretable answer, Borealis 2 alone was outdated on 10 items and Gemma 4 alone on 11 (d = -0.2 percentage po
benchmark - arxiv:2611.02333 · cs.LGWhen Does Adaptive Halting Win? Selection Rules Matter in Looped Language ModelsMahesh Reddy Pagadala
Looped language models can stop iterating early, and recent work shows that post-hoc readouts of the loop trajectory often rival the learned exit gate of the looped model Ouro. Such comparisons reduce each method to one operating point chosen by a selection rule. We ask whether that rule decides the outcome. On a frozen Ouro-1.4B with four loops, we compared a trained stability probe ($π_1$-x) with Ouro's native gate, adapted to a per-question stopping decision, in pre-registered studies; the confirmatory study was designed after an exploratory finding on development data and evaluated once on
benchmark - arxiv:2611.02332 · cs.AILEANGUARD: Lean-Verified Guardrails for Tool-Using AgentsSreehari Sankar, Debargha Ganguly, Vikash Singh, Mohsen Hariri +16
Tool-using language-model agents often execute actions that their domain policies forbid. We present LeanGuard, a runtime that checks every proposed tool call against a policy written in Lean4 before the call reaches the environment. Policies combine state predicates with scoped past-time temporal operators over the execution history, evaluation is fail-closed, and the decision procedure is proved equivalent to a LeanLTL semantics. A denied call can be corrected in two ways. With feedback, the agent receives the violated rule and a hint; with proof-guided repair, Lean metaprogramming searches
agent - arxiv:2611.02319 · cs.AIAccessed, never disclosed: Silent access to protected health information by medical artificial intelligence agentsJae Wang Lee, Abhirup Banerjee, Yeong Chan Lee
Medical artificial intelligence agents should retrieve patient identifiers only when a task requires them, the data-minimization principle behind the HIPAA minimum necessary standard for protected health information (PHI). An agent can retrieve identifiers without revealing them. Such silent access remains invisible to evaluations that inspect only the output. We embedded canaries in synthetic patient records and linked each tool invocation to canary exposure in the final answers across five language models, four boundary levels, and seven attack frames for identifier-free clinical tasks. Thre
agent - arxiv:2611.02314 · cs.CLGRIPS: An Empirical Study of Pre-SQL Grounding Information for SQL GenerationEkata Mitra, Yufei Tao, Kamalamalini Nagasundaram, Ameeta Agrawal
Text-to-SQL systems answer natural language questions over structured data, but can fail before SQL generation by linking question mentions to the wrong table, column, value, or relation. Even valid SQL can then return an incorrect answer. This raises a central question: What pre-SQL information helps a SQL generator, and when does additional information cease to help? We investigate this question on TempTabQA-C and BIRD. Among the audited SQL Static failures, grounding errors form the largest category on both benchmarks. We introduce GRIPS (Grounding Representation and Input Processing for SQ
benchmark - arxiv:2611.02300 · cs.MATrustworthy Execution of Industrial AgentsZhongjie Wang, Mingyi Liu, Xiang He, Xiao Wang +2
Industrial agents increasingly modify persistent engineering objects rather than merely generate answers. Authentication and successful calls cannot establish whether intended changes occurred, evidence supports a conclusion, or historical results remain applicable. We present a framework for trustworthy industrial-agent execution within a single enterprise governance domain. Eight control domains assign responsibilities across authority, planning, composition, tool execution, actual effects, evidence, baseline and use governance, and recovery and evolution. A four-plane architecture and three
multi-agent - arxiv:2611.02286 · cs.AIToward Agentic AI-Ready Workflows in Built Environment Organizations: Knowledge Structures and Interoperability ChallengesAhmet Bahaddin Ersoz, Justin McHenry, Yevgeniya Kovalchuk, Edward Barrett +1
The rise of agentic artificial intelligence (AI) in the built environment is shifting attention from isolated AI tools toward autonomous systems that reason, plan, and act across project and organizational workflows. The effectiveness of such systems depends not only on large language model capabilities, but also on the readiness of heterogeneous data environments across the Architecture, Engineering, and Construction (AEC) sector. These environments combine structured sources, such as CAD and BIM models, schedules, cost databases, and asset registers, with unstructured ones, including technic
retrieval-augmentedai agentagentic - arxiv:2611.02284 · cs.CVSVD-Prune: Spectral Vision Token Pruning under Extreme CompressionYvon Apedo, Martyna Poreba, Michal Szczepanski, Samia Bouchafa
Vision-Language Models (VLMs) and Vision-Language-Action (VLA) models process hundreds of visual tokens, making visual representations a major computational bottleneck. We argue that visual token pruning should be formulated not as identifying individually important tokens, but as preserving the dominant low-rank spectral structure of the visual representation. Based on this formulation, we introduce SVD-Prune, a training-free framework that leverages Singular Value Decomposition (SVD) and statistical leverage scores for globally informed token selection. SVD-Prune is architecture-agnostic and
vision-language-actionvlabenchmark - arxiv:2611.02275 · cs.AIData-Scarce, Multi-Granular, Multimodal: Benchmark Design for Biosynthetic Gene ClustersHanlin Xiao, Eriko Takano, Mauricio A. Álvarez, Rainer Breitling
Biosynthetic gene clusters (BGCs) are co-located genes that encode the biosynthesis of specialized metabolites, an important source of antibiotics, anticancer agents, immunosuppressants, and agricultural bioactives. Deep learning, particularly self-supervised foundation models, has spurred interest in BGC sequence modeling, but evaluation has lagged behind: existing benchmarks often rely on coarse classification over small curated datasets, providing limited discrimination among models and insight into downstream utility. We propose a benchmark-design framework in which an informative benchmar
benchmark - arxiv:2611.02273 · cs.CLWhen the Text Is Silent: Memory Shortcuts and the Validity of LLM Annotation under Evidence InsufficiencyWenchang Wang, Yue Song
Large language models (LLMs) are increasingly used to annotate text for social science research, content moderation, and training data. Their broad knowledge of real-world entities raises a concern: when labeling a text, do they rely on its evidence or substitute what they already know? We call this behavior a memory shortcut and examine when it emerges, what amplifies it, and how to mitigate it. Using 995 widely recognized public figures and four matched control names per figure, we construct counterfactual texts that separate textual evidence from prior knowledge, and evaluate three commerci
memory - arxiv:2611.02270 · cs.LGHinted Scaling LawsSuze van Adrichem, Luke Bailey, Rohith Kuditipudi, Christine Ye +5
Scaling laws predict the outcome of large, expensive training runs by extrapolating trends from smaller, cheaper runs. On downstream accuracy evaluations, these predictions fail when the weaker models produced by the cheaper runs score near zero accuracy, because these scores give no trend to extrapolate to larger scales. We introduce hinted scaling laws, which use privileged information to turn a benchmark into a continuum of easier versions so that even the small models score above zero. We find that accuracy rises smoothly and predictably with the hint level, enabling new types of scaling l
benchmark - arxiv:2611.02269 · cs.AIQubrio: High-Performance Quantum Compilation via Multi-Agent LLM CollaborationJixuan Ruan, Zhuo Cui, Zhengding Hu, Zhongkai Yu +9
Large Language Models (LLMs) offer an adaptable alternative to traditional quantum compilation heuristics, which bottleneck progress by demanding costly manual redesigns whenever hardware evolves. However, naive single-pass LLM approaches fail under the unique challenges introduced by quantum compilation. We introduce Qubrio, a practical agentic framework that significantly reduces heuristic redesigns through a three-level decomposition. By (1) partitioning programs into sequential operation stages, (2) deploying specialized agents for placement, routing and optimization, and (3) using disaggr
multi-agentagentic - arxiv:2611.02260 · cs.MATowards a Theory for the Design of Rules in Multi-Agent Systems: the case of Homogeneous FleetsMatteo Penlington, Lorenzo Fiaschi, Raj Kiriti Velicheti, Andrea Censi +1
The regulation of systems composed of many interacting decentralised agents is a long-standing problem. This problem has typically been resolved by the imposition of rules. Rules resolve the limitations of sub-optimal agents whilst simultaneously aligning the behaviour of the agents with the demands of society. Traffic laws, for example, let a driver cross a junction without seeing around the corner, and transport protocols such as TCP let applications share a network without knowing of one another. The problem is more pressing today, as autonomous systems are deployed in society. In this work
multi-agentagent system - arxiv:2611.02258 · cs.AIBeyond Binary Detection: Type-Aware Hallucination Analysis and Hierarchical Detection in Financial Question AnsweringXiongyu Chen, Sizhen Zhu, Zhuoer Wang
Large language models are increasingly used to answer questions over financial filings, where an unfaithful answer can mislead investment, compliance, or client decisions. Hallucination detectors are a common safeguard, but they are usually evaluated with a single binary score that reveals neither which kinds of errors analysesed nor why. This paper studies hallucination detection on PHANTOM, a long-context benchmark built from SEC filings, at the level of hallucination type. We first analyze 15,443 paired hallucinated and faithful answers without any detector and find that hallucinated answer
long-contextbenchmark - arxiv:2611.02255 · cs.LGMulticlass Classification with Rare Useful Features: Fundamental Limits and the Optimality of Diversity Pursuit Higher Criticism ThresholdingGabriel Moryoussef, Alon Kipnis
We consider classification in high-dimensional feature spaces where only a vanishing fraction of features are useful and the number of classes may be large. While this rare-and-weak useful-feature regime is well understood in the binary setting, its multiclass version raises new questions: a feature can be useful for class discrimination without separating any specific pair of classes. We focus on a rare feature diversity model, in which most features are useless, but a small fraction promote discrimination among classes in a way that can be detected by a per-feature diversity test, such as on
benchmark - arxiv:2611.02240 · cs.ROWhen to Probe and When to Act: Clarity-Aware Risk-Adaptive Control for Interactive DrivingYouran Wang, Devodita Chakravarty, Yiwei Lyu
Safe navigation in interactive traffic requires autonomous vehicles to reason about unobservable human characteristics and the reciprocal nature of multi-agent interactions. Traditional architectures often decouple prediction from planning, leading to the "Frozen Robot Problem" or unsafe maneuvers under high ambiguity because they treat risk assessment and information acquisition as independent objectives. We argue that the notion of safety should not remain static but must instead dynamically co-evolve with the system's internal confidence: the robot should actively clarify ambiguity when nec
multi-agent - arxiv:2611.02237 · cs.LGNonreciprocal Spherical Attention and Retrieval: Sharp Coarsening Clocks and Certified Score CorrectionFaruk Alpay, Taylan Alpay
We develop nonreciprocal spherical attention as a mathematical framework for coarsening and compressed nearest-neighbor retrieval. For two shrinking acute blocks whose centers have zero skew pairing and whose total mass ratio is subexponential in inverse temperature, we prove $\lim_{β\to\infty}β^{-1}\log T_β=1-\cosφ$, without restricting particle-count growth. An explicit three-body flow has two rigorously separated metastable exponents. For finite compression, mass centering yields a sharp $O(r^2)$ attention-output error bound at fixed temperature, with an explicit spherical-normalization cor
memorybenchmark - arxiv:2611.02236 · cs.LGIterative Gumbel Planning for Continuous ControlShaohuai Liu, Weirui Ye, Yilun Du, Le Xie
Search-based planning remains a powerful policy improvement paradigm in reinforcement learning, driving milestone successes in discrete domains. However, extending this paradigm to continuous control exposes a structural mismatch: search evaluates a finite set of action candidates, whereas continuous RL typically relies on Gaussian policies. We show that winner-action distillation into a continuous Gaussian induces a \textit{projection error} and a stepwise variance-expanding gradient tendency when the selected action is sufficiently far from the current mean. Empirically, this tendency is ass
humanoid - arxiv:2611.02230 · cs.LGWho Should Score the Cache? Mid-Depth Layers Are the Best Judges of KV-Cache EvictionEkaterina Seregina
Key-value (KV) cache eviction methods rank tokens by attention scores, and each layer scores its own cache. To our knowledge, no eviction method asks which layer should score. We find, across twenty-one models, that the depth of the scoring layer governs eviction quality. On the eighteen models with per-layer caches, spanning 0.5B to 109B, long-context retrieval accuracy collapses at shallow layers, peaks between 0.4 and 0.75 of depth, and degrades at the last layer. The curve arises because a layer judges tokens well only once it has used the question to find the answer, which takes several l
memorylong-context - arxiv:2611.02222 · eess.SYA Data-Driven, Gain-Scheduled Equivalent Circuit for the B-H Behavior of Power FerritesWilhelm Kirchgässner
An equivalent circuit for the dynamic B-H behavior of power ferrites is identified from measured transients. It consists of classical elements of hysteresis modeling: a nonlinear inductor for the reversible curve, six viscous play branches (inductors in series with diode-bridge sliders) that form a Prandtl-Ishlinskii model, two R-L branches for the magnetic after-effect, a mean-field feedback of H as in the Jiles-Atherton model, and a measurement path. All elements are positive apart from the feedback sources, and the circuit can run in SPICE. Discretized at the sampling step, the circuit is a
memory - arxiv:2611.02215 · cs.LGDiscrimination, Calibration, and Ranking: A Unified Matrix FrameworkRonaldo Cristiano Prati, Farangis Johari
Discrimination, calibration, and ranking are the three pillars of classifier evaluation, yet no single algebraic language spans all three. A recent indicator-matrix account unifies discrimination measures (precision, recall, F1, Matthews correlation) across binary, multiclass, multilabel and related settings, using Boolean cell operations on indicator label matrices, generalised to soft labels by t-norms. Calibration measures (Brier score, log-loss) and ranking measures (nDCG, AUC) fall outside it. We close the gap with two further classes of cell operators: deviation operators phi(y - yhat),
benchmark - arxiv:2611.02211 · cs.CVForecasting Earth Surface Change in Embedding SpaceJingtao Li, Sadiq Jaffer, Anil Madhavapeddy, Pedro Sousa +6
Repeated satellite imagery permits understanding of how our planet changes over time, letting us predict its future and plan mitigating actions. Most existing studies that rely on a time series of satellite observations forecast the spectral values at each pixel for a future target date. However, clouds, shadows, and sensor noise often corrupt these observations, so that spectral forecasters waste much of their capacity on low-level features or noise rather than real change. With the advent of Earth foundation models, we argue that forecasting in embedding space is a feasible and more effectiv
benchmark - arxiv:2611.02208 · cs.LGTest-Time Scaling for Open-Ended Reasoning at Small Model ScaleAdam Y. Ahmed, Abdullah G. Elafifi
A vote over a model's samples can only help when the samples disagree. For open-ended generation the binding question is therefore not how to combine samples but whether a model produces samples worth combining, a question test-time scaling has answered for math but not for free-form text. We study it at small scale with semantic cluster voting, a judge-free aggregator that groups a small instruct model's samples by meaning and returns the largest cluster's representative, with no reward model or self-judging. On a 250-prompt benchmark with three tiers of answer checkability, the same pattern
benchmark - arxiv:2611.02207 · cs.LGBuilt on Broken Assumptions (BOBA): Why Coding Agents Cannot Be Left AloneAlejandro Cuadron, Arpit Gupta
Coding agents are autonomous developers. Given a request, they change the code and report back when done. As such, they can handle not only independent requests but whole chains of them, each request built on the last. Agents are known to report wrong work as done, yet benchmarks still test them one request at a time, on human-written code, checked once. We build BOBA-Bench, 280 chains of 20 consecutive requests drawn from 140 real open-source projects, and evaluate ten coding agents on both. All ten do far worse on chains. On average, they complete 95% of requests taken one at a time, but whe
agentbenchmark - arxiv:2611.02203 · cs.CVDepth Anyviewpoint? Evaluating Monocular Depth Over Viewpoint SequencesTheo Rode, Rui Zhang, Ryan Aubrey, Olivia Li +4
Monocular depth estimation aims to predict depth maps from single RGB images (forgoing the need for stereo camera pairs) and has seen impressive leaps in performance in recent work due to the introduction of affine-invariant losses and the use of foundation model features. These monocular depth models predict affine-invariant depth, which describes the relative positions of objects in the scene instead of absolute depth. While these models demonstrate strong performance on single image benchmarks and qualitatively consistent predictions in video, their effectiveness at preserving consistent sp
benchmarkevaluation protocol - arxiv:2611.02189 · cs.RODirectional Bounded-Confidence Matrix Alignment in Vicsek FlockingHossein B. Jond
This paper proposes a directional bounded-confidence variant of the discrete-time Vicsek multiagent model, in which pairwise heading alignment is modulated by a state-dependent, positive semidefinite matrix that independently gates orthogonal radial and tangential components in the instantaneous line-of-sight frame. We analyze the normalized, matrix-weighted system dynamics, proving a uniform non-vanishing bound on proposal vectors that guarantees global well-posedness across arbitrary network states. For local small-spread states under complete metric interaction, we prove linear geometric co
benchmark - arxiv:2611.02181 · cs.ROHuman Sensorimotor Model Adaptation to Account for Robot Accuracy and Speed Augmentation in Human-Robot Co-ManipulationFrits van Hall, Kevin Haninger, Luka Peternel
In human-robot co-manipulation, the robot's trajectory should accommodate the human's, where human sensorimotor behavior is affected by the task's spatial and temporal constraints, as well as personal factors. Several robot trajectory planning methods have been proposed to embed human sensorimotor models in the robot trajectory planning, but these methods do not consider temporal constraints that adapt to how the human behaves when interacting with the robot. To better align human and robot collaborative motion, this work adapts a stochastic optimal control problem with a final state constrain
manipulation - arxiv:2611.02180 · cs.LGInner-Speech Decoding from EEG Across Three Corpora Under a Fixed Architecture and ProtocolMd Wahiduzzaman Suva, Md Imtiaj Alam Sajin, Esm E Moula Chowdhury Abha, Susham Moula Choudhury Akash +4
Inner-speech decoding from electroencephalography could support non-invasive communication for people with severe motor impairment, but comparison across studies is hindered by dataset- and protocol-specific evaluation. We present a controlled cross-dataset study using one preprocessing pipeline, one within-subject five-fold cross-validation protocol, and a compact dual-attention Transformer (VC-DATNet; approximately 131,000 parameters) across three public corpora: Thinking Out Loud (10 participants), BISpeech (4), and SyncSpeech (3). VC-DATNet achieved per-participant sensor/source hybrid acc
benchmark - arxiv:2611.02177 · cs.LGANCHOR: KL-Anchored Adaptation for Stable Offline-to-Online RLGeonwoo Cho, Jaegyun Im, Doyoon Kim, Lexin Li
Offline-to-online reinforcement learning (RL) follows the pretrain-finetune paradigm: an agent is first pretrained on historical data and then adapted through online interaction. We study offline-to-online adaptation without offline replay, where the agent must improve using only the pretrained model and newly collected experience. A key challenge is to avoid catastrophic failure, a sharp early performance collapse during online fine-tuning, while still achieving strong asymptotic performance. Through controlled experiments, we show that removing offline replay can induce catastrophic failure,
frankaagent - arxiv:2611.02174 · cs.LGBeyond Unit-Circle Eigenvalues: Invariant Bases for Stable State Space DynamicsJiaqian Zhu, Yang Zhang, Junhua Ding, Xiaowei Yu
Diagonal state space models process long sequences at linear cost, yet the same transition is applied at every step of a rollout, so an unconstrained recurrence holds no quantity fixed and the state decays or grows without bound. Current designs constrain one quantity, and preserving one does not guarantee preservation of the others, so a model can hold its norm to machine precision while momentum drifts by hundreds of percent. We claim that the conserved quantities of a diagonal recurrence are layered, that each layer requires a strictly stronger condition on the operator, and that a unit-mod
benchmark - arxiv:2611.02171 · cs.ROGoalDex: Learning Image-Goal-Conditioned Policies for Dexterous In-Hand ManipulationArjun Bhardwaj, Mayank Mittal, Maximum Wilder-Smith, Vaishakh Patil +1
Dexterous in-hand manipulation is commonly formulated around explicit object pose estimation, yet reliable pose tracking is challenging under severe hand-object occlusions and often requires object-specific supervision. This dependence creates a bottleneck between perception and control that can limit robustness and scalability. We present an end-to-end visuomotor framework for dexterous in-hand reorientation that operates on RGB observations and proprioception without intermediate pose estimation. Our approach jointly learns visual representations and control policies through behavior distill
manipulationdexteroussim-to-real - arxiv:2611.02162 · cs.LGRewarded for Not Getting Caught: Why Alignment Training Selects for Conditional Compliance and Cannot Instill NormsKevin Baum, Ruta Binkytė, Gerrit Großmann, Felix Jahn +1
Current alignment training cannot be expected to deliver what it is meant to deliver: AI agents that comply with norms whatever the gains and whether or not anyone is watching. Training that rewards and penalizes behavior can teach norms only through that reward, and a reward can encode a norm only as a price: "do not do X" becomes "doing X costs something", and training charges the cost only when X is noticed. A price, unlike a norm, can be paid: a large enough gain outweighs it, noticed or not. And because the cost is charged only where violations are noticed, training cannot favor a policy
ai agent - arxiv:2611.02153 · cs.CLMulti-Agent Collusion and the J-SpaceMarco Molinari, Gabriele Bettineschi
LLM agents are increasingly deployed in Multi-Agent Systems (MAS), and they introduce the risk of coordination among agents that might escape human oversight. The J-space has been used to uncover deception in single LLMs, but its use for uncovering and steering collusion between agents remains unexplored. We show, with interpretability experiments on DeepSeek-V4-Flash (284B) on the NARCBench collusion benchmark, that colluding agents' J-space holds the collusion itself: before replying, 86% of colluders represent misrepresentation, concealment or remorse. In a steganographic card-counting task
llm agentmulti-agentagent systembenchmark - arxiv:2611.02145 · cs.LGGatedV: Self-Gated Values for Transformer Language ModelsLi-Ming Yu
Standard attention forms query (Q), key (K), and value (V) matrices through learned linear projections of input hidden states. Architectures such as GAU and Mega incorporate nonlinear transformations into attention. In this paper, we propose GatedV, a simple attention variant that applies SiLU only to V before aggregation without introducing additional trainable parameters. We evaluate GatedV against a standard decoder-only Transformer using matched parameter counts, paired initializations, and equal training-token budgets. At 2.33M parameters, GatedV reduces mean held-out perplexity from 59.1
benchmark - arxiv:2611.02139 · cs.LGAdversarial Robustness of Retrieval-Based Fact-Checking: Detecting and Mitigating Evidence ContaminationAlamgir Munir Qazi, John P. McCrae, Jamal Abdul Nasir
Retrieval-based fact-checking systems ground verification in large document collections, creating a critical vulnerability: adversarial evidence contamination. We demonstrate that factually inconsistent but contextually coherent evidence generated to satisfy both retrieval-similarity and linguistic-fluency constraints can systematically mislead verification models. Contaminating just 3.1-7.4% of training corpora with five label-specific strategies causes adversarial content to penetrate up to 35.2% of top-k retrievals, degrading verification accuracy by up to 46.9% relative F1. Existing retrie
benchmark - arxiv:2611.02134 · cs.CLHaloGuard 2.0: An Open-Weights Constitutional Output Guard for Multilingual Response SafetyAshmiya Lenin, Navaneeth Sangameswaran, Preetham S
An input guard decides whether a request reaches the model, but never sees what the model says. Harm that emerges during generation, because a jailbreak passed the input layer or a benign request drifted into operational detail, is visible only on the output side and only as tokens arrive. We present HaloGuard 2.0, an open-weight streaming output guard that assigns a safe/unsafe probability to every response token in a single forward pass, so a deployment can stop the stream once harm is confirmed. The guard reuses the constitution of HaloGuard 1.0, our input guard, to organise a 478,532-respo
benchmark - arxiv:2611.02132 · cs.LGTSSDA: An Autonomous Agent for Constructing a Time-Series Text-to-SQL BenchmarkZhihan Zheng, Shaoxu Song, Haitao Yuan, Minxiao Chen
As agents increasingly undertake data querying, Text-to-SQL serves as a foundational capability for database access. However, for text-to-SQL in time-series databases, high-quality benchmarks and large-scale training corpora remain scarce, limiting the systematic evaluation and domain adaptation of existing models and agents. Existing automated SQL generation approaches lack sufficient autonomy, leaving key challenges unresolved. First, numerous combinations of temporal syntax and functions complicate comprehensive coverage of valid queries. Second, as the corpus grows, generating valuable and
memoryagentautonomous agentagent frameworkself-improvementbenchmark - arxiv:2611.02119 · cs.LGTesting concept-level audits of melanoma classifiers with diffusion counterfactuals and matched positive controlsMd. Shohanoor Rahman, Md. Hasibul Islam, Md Kishor Morol, Moinul Hossain +2
Concept-based explanations of melanoma classifiers are rarely tested against controls with known targets. We built a counterfactual audit from a diffusion model conditioned on seven 7-point checklist concepts and tested it with such controls. Colour-calibrated start noise reduced colour outliers from 78.5-84.8\% to 6.6-13.3\%. Four of seven knobs were target-dominant but also changed other concepts. Quartile-based positive controls were confounded, and the audit failed its pre-specified rule. It recovered the targets of all 16 controls matched on the generator's pseudo-labels and of 12 of 12 c
evaluator - arxiv:2611.02114 · cs.LGCNN-Based Retinal OCT Classification: Research Gap Analysis and a From-Scratch Baseline, ED126-AShkelqim Sherifi
Automated classification of retinal pathologies from Optical Coherence Tomography (OCT) images remains challenging because diagnostically relevant morphological patterns may exhibit substantial inter-class similarity. Although high classification accuracies have been reported on benchmark OCT datasets, cross-study comparison is limited by heterogeneous evaluation protocols and by the limited characterization of compact convolutional neural network (CNN) baselines trained from random initialization. This study addresses this gap through two contributions. First, it provides a focused review of
benchmarkevaluation protocol - arxiv:2611.02107 · cs.AIZT-DAC: ML-Driven Dynamic Access Control Policy for Zero Trust in DER NetworksArun B. Chandrasekaran, Damilola Popoola, Souradeep Bhattacharya, Manimaran Govindarasu
The increasing penetration of Distributed Energy Resources (DERs) in the smart grid has significantly expanded the cyber-attack surface. Traditionally, DER networks rely on static role-based access control (RBAC) mechanisms that assume persistent trust after authentication. This limitation enables adversaries to maintain unauthorized privileges and move laterally across DER systems, disrupting grid operations. Although attribute-based access control (ABAC) enables fine-grained and context-aware access decisions, it does not fully eliminate implicit trust. Zero Trust (ZT) principles can be inco
agent - arxiv:2611.02104 · cs.AIConsolidating Reasoning with Test-Time TrainingHaotian Wu, Zeming Chen, Hao Zhao, Antoine Bosselut
Effective multi-agent collaboration requires finalizing solutions from reasoning that is distributed across the individual attempts of many agents. However, useful intermediate reasoning can remain unexploited when collaboration is limited to voting on conclusions reached among individual agents, or aggregating long, noisy collections of reasoning trajectories in context. In this work, we propose TTT-Consolidate, a framework that uses multi-agent reasoning trajectories as adaptation data to improve a model's subsequent reasoning. TTT-Consolidate jointly consolidates trajectories into parametri
multi-agentbenchmark - arxiv:2611.02102 · cs.LGNorthCUA Technical Report: Turbocharging Computer-Use Agents through Experience at ScaleChaoyun Zhang, Bingxuan Li, Pinxue Guo, Yangfan Gao +26
Scaling computer-use agents from individual actions to complex software work requires experience that connects decisions to complete outcomes. We present NorthCUA, an integrated system for turbocharging computer-use agents through experience at scale. Its Experience Flywheel supplies diverse executable tasks to the NorthCUA Harness, which sustains long interactions, records decisions, and checks task outcomes. The Flywheel uses these records to curate verified trajectories. For unsolved tasks, recursive self-improvement evolves reusable skills and preserves verified milestones so later attempt
self-improvement - arxiv:2611.02101 · cs.ROWhen Delay Meets Drift: Physical Safety of Robot Foundation Models Is a Cross-Layer ProblemLiangkai Liu, Bimal Thapa Magar
Robot foundation models now act in a closed loop, yet their safety is evaluated one variation at a time: robustness benchmarks perturb the scene while assuming instantaneous inference, and real-time execution methods study inference delay in nominal scenes. We argue that physical safety is not a property of the model alone, but emerges from how variations in the physical environment, the model, the computing platform, and execution interact in closed loop. We formalize this as cross-layer variation propagation across these system layers and use an interaction contrast to measure when joint var
vision-language-actionpi0robot foundation modelliberobenchmark - arxiv:2611.02090 · cs.ROLocal Solution Repair for Multi-Agent Pickup and DeliveryEthan Schneider, Jingkai Chen, Seth Hutchinson, Sonia Chernova
Multi-robot systems in automated warehouses must efficiently attend to a constant stream of pickup-and-delivery tasks while avoiding collisions. Existing Multi-Agent Pickup-and-Delivery (MAPD) solutions generate plans that are globally feasible but often locally inefficient. In large-scale, long-horizon deployments, these inefficiencies, such as robots backtracking from target goal locations and idling in congested areas, accumulate over time and degrade performance. This paper contributes Local Solution Repair (LSR), a goal replanning approach that detects suboptimal behaviors at run-time and
multi-agent - arxiv:2611.02084 · cs.LGBlack Hole Love Numbers with Physics-Informed Neural NetworksMaria J. Rodriguez, Luis Fernando Temoche, Pavlos Protopapas
We develop and validate a physics-informed neural network (PINN) framework for computing the tidal Love numbers of black holes directly from their perturbation equations. We consider static perturbations of the Kerr black hole, whose exact hypergeometric solutions provide a controlled benchmark for both the numerical solver and the extraction of the physical response. The framework combines exact enforcement of the horizon boundary conditions with an $H^2$-Sobolev loss that incorporates derivatives of the differential-equation residual. The Love number is then obtained by extracting the asympt
benchmark - arxiv:2611.02079 · cs.LGParetoSaddler: Pareto-Guided Optimization of Agent Harnesses for Quality and CostXingzhi Qian, Sungho Park, Rongyuan Tan, Jianbo Yu +12
Harness optimization can substantially improve LLM agents without modifying model weights, but existing methods primarily optimize response quality. Extending them to joint quality--cost optimization introduces a challenge: applying the deployment preference during intermediate updates can prematurely reject candidates that are valuable for later improvements. We introduce ParetoSaddler, which addresses this mismatch by separating intermediate candidate promotion from final deployment selection. At each iteration, an optimizer updates a parent harness to produce a candidate, and then compares
agentllm agent - arxiv:2611.02076 · cs.ROScene-Adaptive Readout and Decoding in Diffusion Driving VLAsSeungjun Yu, Jiho Choi, Inha Kang, Hojun Choi +2
Diffusion driving vision-language-action models repeatedly execute the backbone at every denoising stage, incurring substantial inference cost, while it remains unclear whether all scenes require the same decoding computation. To distinguish when a trajectory becomes readable from how much further refinement it requires, we analyze trajectory readout across network depth and decoding progress. We identify a depth-decoding dissociation in which late-layer representations already support useful trajectory readout after the first forward sweep, whereas early-layer representations benefit more fro
vision-language-actionbenchmark - arxiv:2611.02075 · cs.AIDecoupled Predictive Decision Making: A Shared Conditional Predictor for AgentsLang Mo, Xingxuan Zhang, Qi Wang, Mingchao Hao +4
Agent decisions depend on empirical relationships that deployment observations reveal and task descriptions alone may leave unresolved. We introduce Decoupled Predictive Decision Making (DPDM), an architecture that formulates heterogeneous agent decisions as conditional prediction with a shared pretrained learner. Language models provide semantic representations of current and historical records, while a general-purpose conditional predictor predicts task-defined targets from the represented evidence. Pretraining supplies reusable inference capability; deployment evidence instantiates problem-
agentbenchmark - arxiv:2611.02074 · cs.AISWE-Trail: History-Guided Trajectory Collection for Coding Agent TrainingZiming Li, Lu Yu, Jiatan Huang, Qing Cui +2
Coding agents solve repository issues through long sequences of exploration, editing, and validation. These trajectories are increasingly used as post-training data, so how they are collected shapes what the trained model learns. A common approach is to sample multiple trajectories independently for each issue and then select or curate them. However, as more trajectories are sampled independently, later trajectories become increasingly redundant, even though earlier trajectories could guide where subsequent collection effort should be spent. A natural way to use this history is to expose it to
agentpost-training - arxiv:2611.02073 · cs.ROBack to the Future: Predicting Long-Horizon Dynamics with Retrospective Correction for Disturbance-Aware RetrievalJiwoong Choi, Haoyu Li, Yiwei Lyu
Retrieving objects from densely packed storage requires anticipating how contact disturbs surrounding objects when limited clearance prevents lifting the target over clutter. However, contact can propagate beyond the target's immediate neighbors, while accumulated prediction errors make learned dynamics unreliable over long planning horizons. We introduce DARE (Disturbance-Aware REtrieval), a geometry-aware relational dynamics framework that uses retrospective correction to improve long-horizon prediction for disturbance-aware retrieval. DARE encodes compact object descriptors and predicts act
action-conditioned - arxiv:2611.02065 · cs.LGFragility Is Not Reducible to Hardness: Conditional Causal Difficulty for LLMsMohammad Reza Rezaei, Dingxian Wang, Nikhil Verma, Sajjad Pakdamansavoji +2
Large language models show large and seemingly arbitrary performance drops when a task is rewritten in ways that leave its requirements untouched. We recast task difficulty as a causal quantity: the treatment effect of meaning-preserving perturbations on a model. A meaning-preserving rewrite can only be observed to hurt a task the model can already solve, so we place the estimand on the principal stratum where the degradation occurs: the conditional causal difficulty (CCD), the probability that a randomly drawn twin breaks a solvable task. Because harder tasks tend to be more phrasing-sensitiv
benchmark - arxiv:2611.02056 · cs.RODo Embodied VLMs Actually Use Their Scene-Graph Memory?Shubh Garg, Divyansh Pandey, Vedant Khandelwal, Niyati Rawal +1
Embodied agents that answer questions about their route need a memory of what they saw, and scene graphs are a natural form for it. A vision-language model (VLM) can receive such a graph as text or through a graph encoder, but accuracy alone cannot show whether the graph holds the episode's facts, reaches the model, or is actually relied on. In this diagnostic study, we introduce a graph-swap test for the last of these: each episode is given another episode's graph, and we count how many answers change. On an embodied question-answering benchmark with routes through two to seven rooms, we use
embodiedmemoryscene graphembodied agentbenchmark - arxiv:2611.02049 · cs.LGReinforcement Learning in High-Energy Physics: A Survey from Accelerator Control to Theory ExplorationLuca Anzalone
Reinforcement learning (RL) is a machine learning paradigm in which an agent learns a decision-making policy by interacting with an environment, guided by a scalar reward rather than explicit supervision. Supervised and unsupervised learning have become indispensable tools in high-energy physics (HEP), but they learn from fixed datasets, whereas many HEP tasks are closed-loop decisions: the data an algorithm observes depend on its choices, the feedback is sparse or delayed, and the objective is often measured on a real machine or in a stochastic simulation, not by a differentiable model. This
sim-to-realagentmulti-agentbenchmark - arxiv:2611.02044 · cs.CVUnified Loss-Aware Density Control for 3D Gaussian SplattingTae-Young Kim, Jiwoo Chung, Jae-Pil Heo
The efficiency and fidelity of 3D Gaussian Splatting critically depend on adaptive density control, which grows and suppresses Gaussian primitives throughout optimization. However, the standard pipeline still relies on heuristic view-space gradients, scale thresholds, and global opacity resets, rather than explicitly assessing the loss-reducing effect of each split, clone, or opacity update. This mismatch can allocate primitives to already-sufficient regions, miss under-reconstructed structures, and destabilize training through blanket opacity changes. In this work, we recast adaptive density
benchmark - arxiv:2611.02041 · cs.LGECGJev: Question-Conditioned Option Scoring for 12-Lead ECG InterpretationZhuoyuan Li, Xidong Wu, Hanyu Cui, Jiansong Gao +3
Clinical interpretation of a 12-lead electrocardiogram (ECG) often requires answering targeted diagnostic questions with clear, quantifiable decisions. We introduce ECGJev, the first ECG framework built on a decision decoder, a pretrained language model trained to return a probability for every listed option. A question-conditioned bridge compresses frozen waveform and spectrogram features into 16 ECG tokens, and a two-stage procedure adapts the bridge and low-rank adapters on the decoder's native option readout using diagnostic labels alone, scoring all options for a question in one forward p
benchmark - arxiv:2611.02037 · cs.LGScalar Smoothness Is One Point on a Besov SpectrumBum Jun Kim, Anastasis Kratsios
Studies of neural network generalization and robustness often summarize smoothness with a single number, such as a Lipschitz constant or a Hessian curvature statistic. However, Lipschitz constants summarize global worst-case sensitivity and Hessian statistics summarize local curvature; neither captures how model responses vary with the size, direction, and location of perturbations encountered in real-world deployment. To measure smoothness under these perturbations, we introduce a finite-difference roughness spectrum that profiles model responses across these conditions. Our roughness spectru
post-training - arxiv:2611.02033 · cs.CLHow Green Is Your Manifesto? Scoring Environmental Positions in the 2024 EU Elections with an LLM PanelFranco Corti, Sajid Siraj
Locating political parties on environmental and climate policy is central to the study of party competition, yet the established instruments, expert surveys and the hand coding of manifestos, are costly, periodic, and slow to extend to new documents. We ask whether a panel of large language models can serve as a cheaper and reproducible alternative. Treating each manifesto as a set of sentences, three models from independent providers score every sentence on two dimensions, environmental protection and climate change, separating a party's position on an issue from the salience it gives it, and
benchmark - arxiv:2611.02028 · cs.AILARK: Lean-Assisted Reasoning for LLM-Generated GPU KernelsZixi Zhang, Jiayi Nie, Haoran Wu, Yujie Yan +6
Efficient GPU kernels are typically low-level and hardware-specific, and LLM agents can now generate them faster than experts can verify. Existing approaches rely on testing, which provides evidence but not a guarantee of correctness. We introduce LARK, an agentic framework that automatically and formally verifies GPU kernels against their PyTorch reference implementations. LARK proves equivalence in functional and memory behavior under exact arithmetic and establishes a new trust boundary: human effort is only needed to verify operator-level Lean semantics ahead of kernel generation time, ins
memoryagentllm agentagentic - arxiv:2611.02026 · cs.CVReconstruction as Spatial Evaluation: Diagnosing Cross-View Spatial Intelligence in Large Vision-Language ModelsJunfei Wu, Qiuyue Wang, Xuhong Huang, Qihang Peng +7
The spatial intelligence of large vision-language models (LVLMs) requires the ability to combine local visual observations into a geometrically consistent representation of the physical world. Many static benchmarks assess isolated spatial judgments, such as relative direction or distance, while interactive embodied evaluations combine perception with exploration, planning, control, and simulator-specific execution. To bridge this gap, we introduce RECONSPATIAL, a benchmark centered on multi-view 3D scene reconstruction across indoor, tabletop, and outdoor environments. Given real multi-view i
embodiedbenchmark - arxiv:2611.02018 · cs.CLCULTURE-CMT: Benchmarking Contextualized Social Media TranslationLinjuan Wu, Dingming Li, Yiwen Qiu, Xinze Lyu +4
Social-media comments are interactions, yet they are often translated and evaluated as isolated text. This mismatch is especially problematic when meaning depends on surrounding discussion, community-specific slang, or culturally situated expressions. We introduce CULTURE-CMT, a Chinese-English benchmark for contextualized social-media comment translation that restores the post and conversational context naturally available to users. Its 572 manually verified examples are organized into Slang, Context, and Both, enabling diagnostic analysis of distinct and interacting sources of translation di
retrieval-augmentedbenchmarkscalable evaluationscalable eval - arxiv:2611.02017 · cs.CLSwitch-SWE: Do Coding Agents Hold Focus When Switching Between Unfinished Tasks?Shuochen Chang, Qianli Ma, Yibo Miao, Qingyang Liu +6
Software development sessions often involve several unfinished tasks rather than one issue carried through to completion. As coding agents take on repository-level development, they need to recover an earlier task's requirements and partial progress after intervening work. Existing repository-level benchmarks typically score one issue-to-patch episode and do not directly measure whether an agent can resume a task after advancing other unfinished work in the same repository. We introduce \textbf{\ourmethod}, a benchmark for focus-task resumption in sessions that interleave unfinished tasks draw
agentbenchmark - arxiv:2611.02011 · cs.CLMedPsy: State-of-the-Art Small Medical Language Models for Efficient Edge DeploymentDavide Vitabile, Alexandro Buffa, Akshay Nambiar, Amril Nazir
Medical language models promise expert clinical support, but the strongest open-source LLMs remain too large for private edge-device deployment, while compact medical models still trail substantially on knowledge-intensive and real-world healthcare tasks. Reasoning models that fit edge memory budgets can produce long reasoning traces, increasing sequential decoding, end-to-end latency, energy consumption, and KV-cache memory pressure. We present MedPsy, a family of text-only 1.7B and 4B medical language models designed for edge deployment, targeting this accuracy-generation-efficiency trade-of
memorypost-trainingbenchmark - arxiv:2611.02009 · cs.CLRetrieval-Enhanced Cultural Evaluation for Social Media Translation TrainingLinjuan Wu, Yiwen Qiu, Xinze Lyu, Dingming Li +10
Social media translation requires preserving meaning while capturing the cultural nuances of user-generated content (UGC). However, gaps in the cultural knowledge of LLM-based evaluators can lead to inaccurate assessments and unreliable training rewards. To improve these signals, we use retrieval-enhanced cultural evaluation to train CUE-MT, a family of social media translation models. A calibrated, domain-specific Judger guides supervised data refinement and is augmented by an Agent for hard-example mining and reinforcement learning. The Agent retrieves relevant external evidence as needed an
agentbenchmarkevaluator - arxiv:2611.02008 · cs.CVNAS-PO: Native Attention-Strategy Policy Optimization for Vision-Language ModelsKe An, Shizhe Xiang, Yue Liu, Liang Chen +4
Reinforcement learning with verifiable rewards (RLVR) has become an effective post-training paradigm for vision-language models (VLMs). Standard RLVR, however, assigns identical outcome-derived credit to all generated tokens regardless of their perceptual relevance. Recent perception-aware methods estimate visual reliance through image perturbation or masking. These indirect proxies require perturbation branches, while corruption sensitivity does not necessarily identify the visual evidence used for reasoning. We instead shape model perception through native attention, an intrinsic model signa
post-trainingbenchmark - arxiv:2611.02003 · cs.LGMIRA: Mutual Information guided calibration for Reliable Test-Time AdaptationHyeongyu Kim, Mingyeong Jang, Youngjun Song, GeonHui Han +1
Test-time adaptation (TTA) updates a deployed model using only unlabeled target data, but standard entropy-based methods often rely on deterministic confidence and can reinforce overconfident errors under distribution shift. We propose MIRA, Mutual Information guided calibration for Reliable Test-Time Adaptation, a source-free online adaptation method that uses lightweight last-layer Gaussian perturbations to probe prediction reliability. For each target batch, MIRA reuses a single feature extraction pass and samples perturbed classifiers around the source classifier, forming an efficient loca
benchmark - arxiv:2611.02001 · cs.AITrackMem: Tracking Evolving Beliefs in Long-Horizon Financial DialogueAustin M. Y. Cheung, Harsh Vardhan Gupta, Go Nam Lui, Lok Kan Chan +2
Memory systems for LLM agents are evaluated mostly on social dialogue, where facts accumulate and rarely change, so remembering reduces largely to finding the turn in which something was said. Financial dialogue breaks this assumption: figures are revised, stances shift without being restated, rules stated once govern later questions, and a single company is discussed under several analytical lenses. Memory systems that retrieve the passages most similar to a question, or the most recent ones, therefore return superseded values, miss rules that share no wording with the question, and cannot re
memoryllm agentbenchmark - arxiv:2611.01996 · cs.AISafety4DM: A Vision-Language Framework for Risk-Aware Dynamic Maps in Vehicle-Road-Cloud Collaborative DrivingZhengshu Zhou, Shaoyu Wang, Lu Tao, Xinhui Hu +2
Dynamic map (DM) is a core information infrastructure for vehicle-road-cloud (VRC) cooperative autonomous driving. It fuses perception data from multiple sources to overcome the inherent line-of-sight and field-of-view limits of single-vehicle perception. However, DM still falls short on driving safety in two ways. First, DM data specifications already define road conditions semantically, but no pipeline yet turns real-world road observations into fine-grained risk entities in the DM. Second, existing DM systems lack a natural language interface for driving risks. To close these gaps, we propo
benchmark - arxiv:2611.01986 · cs.AIRubricWorld: World Model-Guided Rubric Optimization for LLM GradersHongji Pu, Hang Li, Yucheng Chu, Jiliang Tang
Rubric optimization improves LLM grading by revising scoring instructions while keeping the grader's parameters fixed. Existing optimizers often judge edits by aggregate validation performance, limiting the supervision they retain about which errors an edit corrects or introduces. Testing each candidate also requires rescoring validation answers, which limits how many edits can be explored within a fixed budget. We propose RubricWorld, a framework that learns from previous edit evaluations to guide further search. Its world model predicts an edit's correction and degradation rates at score bou
world model - arxiv:2611.01982 · cs.CVSurgStream: A Workflow-Anchored Proactive Streaming Framework and Benchmark for Real-Time Surgical Video UnderstandingRuipeng Ma, Qing Xu, Jie Liang, Zhen Chen
Surgical video understanding is essential for intraoperative assistance such as workflow monitoring, instrument preparation and procedural documentation. Delivering such assistance in real time requires a model to jointly perceive the ongoing scene, remember past surgical events, and respond before clinically relevant moments pass. Existing approaches provide limited support for jointly retaining procedural evidence and deciding when to respond in streaming surgical video. In this work, we propose SurgStream, a workflow-anchored proactive streaming framework for real-time surgical video unders
memorybenchmark - arxiv:2611.01976 · cs.AIBeyond Prompt Injection: The Tool Update Gap in LLM Agent SecurityShiyu Kan, Yue Qi, Yufei Han, Xinjian Luo +2
Large language model based agents increasingly rely on external tools to invoke services and interact with web resources. These tools can evolve through software updates, yet agentic frameworks may continue trusting them without re-verifying their safety after implementation changes. This tool update gap creates a supply-chain vulnerability in agentic systems: because updated tool code executes outside the LLM's context window, malicious functionality embedded in tool updates can bypass defenses centered on model-visible interactions, leaving agents exposed to stealthy, persistent threats that
agentllm agentagentic - arxiv:2611.01970 · cs.CVKnowledge-Guided Floodwater Level Reasoning for Cloud-Edge Collaborative MonitoringZhengshu Zhou, Shaoyu Wang, Ziyi Geng, Lu Tao +1
Frequent urban floodwater poses a serious threat to the safety and efficiency of intelligent transportation systems. Existing vision-based methods for floodwater level detection face two main limitations. First, the lack of explicit reasoning leads to poor detection performance and little interpretability. Second, these methods struggle to generalize to the complex and varied real-world road scenes seen from different viewpoints. To address this, we reformulate the task as an explicit visual reasoning problem and propose HydroReasoner, a floodwater level reasoning framework designed for cloud-
memoryknowledge graphbenchmark - arxiv:2611.01966 · cs.AILaunder-Bench: Can Repository-Level Agentic Code Laundering Be Detected?Yuntong Hu, Hongji Pu, Zimo Li, Ruomeng Li +4
Coding agents can now rewrite repository-scale codebases while preserving tested behavior, creating a new challenge for plagiarism detection, license auditing, and software provenance. We ask whether such rewrites can still be traced to their source. Existing clone benchmarks emphasize fragments and bounded edits, while repository-transformation benchmarks evaluate functional correctness rather than provenance. Moreover, ranking the most similar source does not establish derivation when independent implementations can be similarly close. We define repository-level agentic code laundering and i
agenticbenchmark - arxiv:2611.01965 · cs.LGOC-OPD: Observation-coordinated Supervision for On-Policy DistillationRuoran Xu, Yilun Zhao, Ziyu Chen, Arman Cohan
Reverse and forward Kullback--Leibler (KL) divergences provide complementary supervision in on-policy distillation: reverse KL refines the student's active token rankings, and forward KL allocates mass to teacher-supported candidates. At student-generated prefixes, we form a four-observation matrix by treating the student and teacher as both candidate sources and evaluators. The diagonal self-observations measure uncertainty, while source-centered cross-observations recover the two KL objectives. We establish an exact identity equating the Fisher inverse-metric coupling of their logit gradient
evaluator - arxiv:2611.01962 · cs.CVObject-Only Is Not Context-Free: A Counterfactual Diagnostic for Context Removal in Object RetrievalBoaz Meivar, Shai Avidan
Object-centric retrieval pipelines often mask the background before pooling backbone features, aiming for a descriptor that reflects the object rather than its scene. Whether this truly yields an object-only descriptor, however, is rarely tested. We introduce the Keyhole Aperture Probe (KAP), a counterfactual diagnostic that repurposes the ground-truth masks of existing panoptic segmentation datasets (COCO, ADE20K) for retrieval: it holds the pooled object region fixed and varies only the pixels outside it, summarizing a backbone's context dependence in a single number. We find that the pooled
evaluation protocol - arxiv:2611.01958 · cs.ROAnswering Without Arriving: An Evidence Audit for Embodied Question AnsweringShubh Garg, Divyansh Pandey, Niyati Rawal, Vedant Khandelwal +1
Embodied question answering asks an agent to navigate, observe, and then answer questions about what it saw, yet answer accuracy alone cannot tell whether any of this happened. A plausible answer can come from the question, the room label, or the answer choices, and a correct answer does not show that the supporting observation was ever made. We present a multi-room benchmark whose 200 test episodes and 800 questions are stratified by route extent, with an evaluation ladder that holds the questions fixed while varying only the evidence supplied to the answerer, from text-only probes through cl
embodiedagentbenchmark - arxiv:2611.01954 · cs.CLTowards Self-Evolving Narrative Systems: Real-Time Adaptation and Long-Term Optimization with LLMsRuisheng Sun, Jiaqi Shao, Weijia Han, Linxi Ju +8
Traditional interactive narrative systems offer strong authorial control and structural coherence but often restrict player agency. In contrast, LLM-driven narratives enable open-ended generation and greater flexibility, yet frequently suffer from inconsistency and diminished narrative control. To navigate this trade-off, we propose a hybrid narrative framework that dynamically balances pre-authored content with real-time LLM generation via semantic similarity alignment. Additionally, sustaining narrative quality over extended gameplay remains a core challenge in LLM-based storytelling. To add
self-evolving - arxiv:2611.01953 · cs.AIStep-Zero Attack: Agent Safety Before ObservabilityTian Qin, Yuhang Jiang, Junzhe Chen, Lit Sin Tan +1
Existing attacks and defenses for coding agents act once content enters the model context: injected text persuades the model, and guards inspect what the model reads. Yet to judge whether processing an unknown archive, network response, or image is safe, an agent often must process it first, and that step can do harm before any content reaches the model. We call this first processing Step Zero, and harm done by it a Step-Zero attack. Across seven harnesses and six models, each agent received only a path or URL to a crafted input, paired with a benign twin showing the same visible fields. Step
agent - arxiv:2611.01952 · cs.AIHippocampal consolidation without replay in fully fine-tuned networksKeuntae Kim, Yong Suk Choi
Networks that keep learning overwrite what they know. Prevailing remedies replay stored data, distil from old models or freeze the backbone. The brain suggests a different division of labour: the hippocampus indexes where experience lies in cortex while the cortex keeps learning. We show that this division lets a fully trainable network keep learning without cross-task replay, a teacher or inference-time memory. After each task, a training-only hippocampal memory indexes the input geometry of every group of synapses sharing an input, residual energy included, and limits each synapse's movement
memory - arxiv:2611.01946 · physics.opticsExperimental demonstration of phase-conjugated twin-wave coherent reception over 100 km with injection locking assisted phase stabilizationZhenlin Zhao, Xingwen Yi, Dawei Wang
Recovering complex optical signals with high sensitivity is central to coherent communication and sensing. Beyond conventional homodyne and heterodyne reception, phase-conjugate twin-wave encoding and phase-sensitive joint detection offer an alternative strategy with enhanced sensitivity. Extending this approach to long-distance transmission of high-order-modulated signals is challenging because optical phase coherence must be maintained at the receiver. Here we experimentally demonstrate injection-locking-assisted phase-conjugated twin-wave reception of 200-MBaud signals with modulation forma
benchmark - arxiv:2611.01943 · cs.AIBeyond Multimodal Emotion Conflict: Temporal Context-Aware Hidden Emotion Detection with Multimodal Large Language ModelsYonghao Wang, Jiaxin Zhou
Multimodal emotion recognition commonly combines textual, acoustic, and visual signals to infer affective states. Recent work has begun to study conflicts among modalities, but a conflict does not necessarily indicate that a person is concealing an emotion. It may also result from mixed emotions, individual expressive styles, or contextual factors. This proposal aims to distinguish ordinary emotional inconsistency from emotion concealment and to infer the underlying emotion when supported by evidence. We plan to construct a multimodal benchmark with explicit annotations of expressed emotion, i
benchmark - arxiv:2611.01939 · cs.CVVLSplat: Vision-Language Guided Object-Centric 3D Gaussian Splatting via Scene GraphYuntae Jeon, Younho Jeon, Sujin Jin, Sungho Jo +1
Learning semantic representations for 3D Gaussian Splatting has recently emerged as a promising direction for 3D scene understanding. However, existing 2D-to-3D semantic lifting approaches rely primarily on 2D segmentation supervision and lack object-level priors, causing the learned representation to be driven by local mask evidence rather than holistic object structure. In this paper, we present VLSplat, a vision-language guided approach for object-centric semantic lifting in 3D Gaussian Splatting. Our method augments mask-based lifting with object-level semantic priors derived from vision-l
scene graph - arxiv:2611.01934 · cs.CVTriPhysBench: Evaluating Physical Intelligence for Embodied AgentsZhongxiang Sun, Chenghao Yang, Haojie Ding, Boheng Zhang +3
Embodied agents need physical intelligence to understand their surroundings through physical perception, physical knowledge retrieval, and physical reasoning. These capabilities must work together to interpret events and make quantitative judgments. Existing benchmarks advance physical understanding and general-purpose tool use, yet the coordination of these capabilities in physical scenarios remains insufficiently characterized. Inspired by Popper's Three Worlds, we introduce TriPhysBench, a benchmark of 410 questions from 200 videos that brings physical perception, physical knowledge retriev
embodiedembodied agenttool usebenchmark - arxiv:2611.01930 · cs.AIRECEIPT: Measuring and Repairing the Return Assumption of LLM Agents in Web Service OperationsBoyang Yang, Ziyao Yang, Yu Zhang, Duo Chai +2
LLM agents operate Web infrastructure from the command line and rely on a return assumption, namely that each tool call returns within a bounded time and with its requested action finished. Many commands, however, ask a question before they act, and the agent's harness either closes the program's input or keeps it open as a terminal at which nobody types. With closed input, the program can return status 0 without the requested action. With open input, it waits until a time limit, and the agent pays in time and tokens. To the best of our knowledge, this paper is the first to measure and repair
agentllm agentbenchmark - arxiv:2611.01927 · cs.AIIs Your Long-Horizon Benchmark Really Deep? Measuring and Synthesizing Depth in Terminal TasksYu Jiang, Yue Huang, Hang Hua, Yuyang Wu +5
Long-horizon terminal tasks are used to evaluate and train agents. Human-curated datasets are scarce, so many synthesis methods have emerged. They usually measure the length of a task by the turns a model needs to pass it, and its difficulty by labels the LLM sets during generation. We find, however, that the turn count largely reflects the habits of the model, not only the demands of the task, and that the labels barely track how often models pass. We describe a terminal task by its work, the number of operations it requires, and its depth, the number of turns that must happen in sequence bec
agentbenchmark - arxiv:2611.01926 · cs.CLTAC-CoT: Tool-Empowered Implicit Chain-of-Thought via Interleaving Explicit TokensJuseon Do, Sangmin Bae, Jeesu Jung, Hwanjun Song
LLM-based agentic systems are increasingly combined with CoT reasoning to solve tasks that require accurate reasoning, external knowledge, and multi-step deliberation. Yet, explicit CoT incurs substantial generation overhead, while implicit CoT, which replaces textual rationales with compact continuous tokens, does not naturally support the explicit communication interface required for agentic tool-use. We propose TAC-CoT, a tool-empowered adaptive implicit CoT framework that replaces explicit reasoning traces with compact latent reasoning segments while retaining textual tokens only for tool
agentictool-usebenchmark - arxiv:2611.01921 · cs.CVVision-language classification of Torsades de Pointes risk from pseudo-coloured ECG images: do T-wave morphology and colour context help or hinder?Dario Labrador Alonso, Alaa Alahmadi
Contextual prompting is widely presented as a cheap route to better performance from vision models, but the evidence comes largely from benchmarks in which an attribute named in a prompt is also present in the image, confounding the two channels. We investigate the significance of context for a semantically complex biomedical signal classification task using a domain-adapted vision-language model, BiomedCLIP. Building on our previous work that renders electrocardiogram (ECG) signals as colour-enhanced images to flag Torsades de Pointes risk using the QT-nomogram and pseudo-colouring technique,
benchmark - arxiv:2611.01918 · cs.AIAgenticMobility: An Open-Source Platform for Multi-AI-Agent Mobility DynamicsJiaheng Xiao, Jiangtao Han, Jiachen Wang, Honghai Li +4
Understanding mobility dynamics and the effects of urban transportation management policies requires examining how individual travel decisions, traffic conditions, and policy interventions influence one another over time. We present AgenticMobility, an open-source experimental platform for studying how mobility AI agents and other AI agents interact, adapt through repeated experience, and collectively shape mobility outcomes. Persistent agents retain identities and relevant states across rounds, with explicit interfaces defining information access, admissible actions, and decision schedules; l
ai agentagentic - arxiv:2611.01900 · cs.CVSystematic Evaluation of Robustness in Fine-tuned ModelsAngela John, Hamayoon Behmanush
Our study systematically evaluates the robustness of fine-tuned pre-trained models across different datasets and attack types, aiming to provide insights for researchers in selecting pre-trained models for image classification tasks while prioritizing robustness against potential attacks. Our work explores the influence of fine-tuning pre-trained models with different datasets and attack types on their robustness and overview of the appropriate pre-trained model for image classification based on factors such as computation resources, memory usage, and accuracy on clean data. We fine-tune the m
memory - arxiv:2611.01898 · cs.LGRobust Remote ECG Monitoring under Energy Constraints: Anomaly Filtering versus Predictive UncertaintyMustafa Fuad Rifet Ibrahim, Maurice Meijer, Alexander Schlaefer, Peer Stelldinger
Machine learning for remote electrocardiogram monitoring must handle noise and diseases absent from training under tight resource constraints. We investigate whether upstream unsupervised anomaly detection can improve monitoring decisions more efficiently than classifier-derived uncertainty. We evaluate five detectors on PTB-XL and BUT QDB and compare a two-stage filter-classifier system with Monte Carlo Dropout and deep ensembles under unseen-disease and noise conditions. Performance is averaged over five training seeds. We measure single-pass energy, latency, and graph memory on a microcontr
memory - arxiv:2611.01878 · cs.CLDon't Guess, Ask Wisely: Benchmarking and Training Proactive Elicitation in Open-Domain LLM ConversationsRongyu Zhang, Mutian Bao, Jinjian Zhang, Yuetian Du +7
When user queries are incomplete, large language models should ask targeted questions before committing to an answer, a capability we call proactive elicitation, yet evaluation on real user-assistant queries across seven LLMs (8B to frontier) shows that fewer than 11% of underspecified queries trigger a follow-up question. To benchmark and train this capability, we introduce VagueBench, where each seed query is annotated with structured intent and detail units then degraded into intent-vague and fully-vague variants via controlled removal of linguistic cues, yielding known hidden information s
benchmark - arxiv:2611.01870 · cs.MAAutomated Stateful Specialization for Adaptive Agent SystemsMyan Vu, Harrish Ayyanar, Jiang Pang, Anwiketh Reddy +1
Current automated agent design frameworks produce either static workflows that lack adaptability or per-query optimizers that prevent the accumulation of deep, agent-level task expertise. We propose a new direction that reconciles these paradigms: creating stateful teams of specialist agents that accumulate knowledge over time and can be reconfigured for novel tasks entirely without human intervention. To this end, we introduce ASpec, a framework that manages this full agent lifecycle by first autonomously discovering specialist archetypes via evolutionary search and then cultivating their exp
agentagent systembenchmark - arxiv:2611.01866 · cs.LGThe (Butter)fly Dream: Linear Separability of World-Model Rollouts in a Connectome-Constrained Visual ModelJunichiro Niimi
Zhuangzi dreamt he was a butterfly and, on waking, could not tell which was the dream. We ask a version of his question of a model of the fly brain: placed inside a world generated by a world model, can a fixed brain tell that the world is a dream? We treat it as a question about what a connectome-constrained model trained on one task makes available beyond that task: do its responses make the defects of a dream linearly decodable? The observer is a published model of the fly visual system trained only for optic-flow estimation; reality is the MPI-Sintel movie on the model's 721-column retina;
world model - arxiv:2611.01860 · cs.CVAdvances in Agentic Computer GraphicsHengyu Liu, Jihyeon Je, Ruiqi Wang, Mingxuan Wu +10
Computer graphics is built on code. Large language model (LLM) agents can now generate graphics code and iteratively refine it through various feedback. This capability opens the door to a potentially new paradigm of computer graphics, with graphics algorithms increasingly built around agentic systems to automate progressively complex tasks. Over the past years, research on agentic graphics has expanded rapidly across a diverse range of tasks. This rapid growth makes it increasingly difficult for manual literature reviews to keep pace with new developments. At the same time, similar ideas are
agentic - arxiv:2611.01858 · cs.CVLearning-to-Memorize: Dynamic Context Management for Long-Horizon Autoregressive Video GenerationHaowei Zhu, Qijie Wang, Jia Li, Xing Wang +6
Autoregressive video generation has achieved impressive results on short clips, yet generating long videos remains challenging due to error accumulation over extended horizons, where each predicted frame depends on potentially imperfect previous frames. Existing approaches typically rely on static context management strategies, such as sliding-window caches that discard early frames or fixed sink frames that permanently preserve initial content. These designs either lose critical long-range context or anchor generation to outdated information, which can lead to degraded fidelity, motion stagna
memory - arxiv:2611.01852 · physics.opticsSpin-induced lateral acoustic forces on chiral and achiral particlesYurou Jia, Shanjun Liang, Wanyue Xiao, Shubo Wang
Acoustic spin enables intriguing chiral sound-matter interactions, yet its mechanical consequences remain much less explored. Here, we demonstrate spin-induced lateral acoustic forces on chiral and achiral particles coupled to a substrate. Under plane-wave incidence, left- and right-handed chiral particles scatter fields with opposite acoustic spin that selectively couple to counterpropagating guided waves, generating lateral forces of opposite signs on the enantiomers. The lateral forces can also be realized for achiral particles by employing acoustic vortex-beam illumination. We experimental
manipulation - arxiv:2611.01844 · cs.ROOpenNWM: A Generalist Navigation World Model with Latent Action PretrainingDujun Nie, Ruiqi Song, Zixuan Wang, Hanqing Wang +5
Navigation world models predict future observations conditioned on navigation actions to support planning, but their generalization across real-world environments remains constrained by the limited coverage of action-labeled data. Action-free navigation videos provide diverse visual experience, yet lack the recorded actions required for action-conditioned training. We introduce OpenNWM, a generalist navigation world model with latent action pretraining. It learns universal latent actions from large-scale action-free videos through visual prediction pretraining, and further aligns them with phy
world modelaction-conditionedpost-trainingbenchmark - arxiv:2611.01843 · cs.AIMEX: The Musical Expectation Benchmark SuiteMathias Rose Bjare, Augustin Bouquillard, Giorgia Cantisani, Seung-Goo Kim +2
We propose the Musical Expectation (MEX) benchmark suite for perceptually validating computational models of musical expectation, i.e., models that estimate music information dynamics (MID) measures of perceived musical expectations, such as surprisal or uncertainty from symbolic music or musical audio. MEX collects and extends tasks for evaluating such models previously proposed in music information retrieval and cognitive sciences using standard datasets, and evaluates whether: (1) proposed MID measures decrease with repetition due to in-context learning of the computational model, (2) peaks
benchmark - arxiv:2611.01840 · cs.AISelf-Evolving Agentic Workflows via Semantic Decomposition of Complex TasksZichuan Fu, Junjia Qi, Xian Wu, Wenlin Zhang +3
Agentic workflows organize complex tasks into reusable procedures. However, invoking an LLM at every step incurs repeated inference costs and delays. We propose a framework for self-evolving agentic workflows through semantic task decomposition. Each subtask has explicit applicability conditions and a procedure that executes code or invokes a model as needed. An agent uses execution feedback to refine the decomposition and update these procedures. Experiments in the GLEE competition show the framework on bargaining, negotiation and persuasion. Most observed final-decision procedures are implem
agentagenticself-evolving - arxiv:2611.01836 · cs.AINEIGHBOR: Aligning Service Behaviour and Evidence Use in an LLM Agent for Community Resident ServicesZongyou Yang, Yinghan Hou, Xiaokun Yang
Residents who contact a community service desk about a repair or a service charge need answers they can act on, with facts that hold for their case and commitments the desk can keep. In tool-augmented LLM agents, such facts come from documents and records, and the conduct of a service answer is usually specified in a system prompt. We present NEIGHBOR, an LLM agent for community resident services that trains this conduct into the model that drives its tools. NEIGHBOR is deployed on a live resident-service platform for a district of an East Asian city of more than ten million people. It adapts
agentllm agent - arxiv:2611.01835 · cs.LGReFrame: Silence-Aware Compute Reclamation for Full-Duplex Speech ModelsZheng Zhou, Guanyi Chen, Jing Wei, Haoyuan Liang +2
Full-duplex speech models execute dense model inference for every audio frame throughout a live interaction. This uniform framewise schedule creates sustained GPU load across dialogue states whose computation has strong, reusable temporal structure. ReFrame formulates full-duplex inference as a silence-aware computation problem. Mutual Silence Speculation (MSS) operates when both participants are silent: it evaluates multi-frame Temporal continuations, verifies text and hidden-state stability, and applies batched Mimi audio-token correction. Listening Frame Folding (LFF) operates while the use
agent - arxiv:2611.01834 · cs.AILegoEnv: Self-Evolving Agentic Environments via Modular RecombinationZhuosong Jiang, Chengxu Wu, Junqi Liu, Yujie Yi +10
LLM agents learn by interacting with environments, yet existing environments are designed for specific tasks and fail to generalize to new ones. Recent methods automate environment construction with LLMs. However, they generate each environment from scratch, neither reusing parts shared across environments nor carrying a fix from one environment to the rest. To address this, we propose a new paradigm of evolving environment atoms, which composes environments from reusable atoms (intent, rule, data and server) and evolves the atoms when an environment fails to give reliable feedback (e.g., an a
llm agentagenticself-evolving - arxiv:2611.01815 · cs.LGWhen Does Latent Prediction Error Support Physical Surprise Claims in Video World Models?Anton Hristov, Aleksandar Slavov, Vasil Vasilev
Prediction errors from video world models are often treated as physical surprise, but they can also reflect appearance, static content, motion, or ordinary prediction difficulty. We present an evaluation framework that separates what frozen representations contain, how they organize appearance and physical differences, and what changes a model's own prediction error. On IntPhys2, a benchmark of physically possible and impossible events, we compare representations from latent prediction, masked pixel reconstruction, and simple visual statistics. Appearance remains easy to decode, while V-JEPA2
world modelv-jepaaction-conditionedbenchmarkevaluation framework - arxiv:2611.01814 · cs.ROControl Barrier Functions for Safety-Aware Latent-Space Motion PlanningChandrakiran Ravula, Siddhanth Ramesh
Latent-space path planning (LSPP) compresses the configuration space of a robotic manipulator into a compact, differentiable manifold learned by a variational autoencoder, enabling fast gradient-based planning without explicit inverse kinematics. Existing LSPP frameworks enforce obstacle avoidance with a binary collision classifier, whose saturated gradients near obstacle boundaries and lack of temporal structure cause the optimizer to stall and provide no forward-invariance guarantee. We replace the classifier with a neural control barrier function (CBF) learned directly in the latent space.
manipulatorfranka - arxiv:2611.01799 · cs.LGLearning AI Agents, Market Makers, and Stress in Automated MarketsIrene Aldridge
Automated markets increasingly route order flow from learning trading agents through automated market makers (AMMs). We model trader archetypes as AI agents defined by their reward functions and analyze the feedback they create on a constant-product AMM. Herd-dominated and arbitrage-dominated markets of equal size have identical return distributions and turnover, so no rule based on volatility, value-at-risk, stress frequency, or volume can tell them apart; yet their stress episodes differ in length by a factor unbounded in the feedback-to-noise ratio $β/s$. Serial statistics identify the feed
ai agent - arxiv:2611.01791 · cs.CVRIFT: Residual-Informed Flow Transfer for One-Step Generative Video CompressionQizhe Wang, Jiaye Fu, Feng Ye, Haocheng Tang +4
Recent advances in video generation have enabled generative video compression to reconstruct details lost during coding using pretrained spatiotemporal priors. However, a substantial gap remains between codec-induced degradation and the perturbation trajectories learned by flow-matching models. In this paper, we identify a compression-dependent directional mismatch that degradation-strength-based timestep calibration alone cannot resolve. Compressed latents remain strongly aligned across codec operating points, while their reconstruction residuals vary substantially in magnitude and direction.
benchmark - arxiv:2611.01783 · cs.CVOCBI: Order-Sensitive Cross-Branch Interaction for Image Manipulation LocalizationZhouli Yu
Image manipulation localization combines semantic structure with forensic traces to identify modified pixels. Within-branch spatial mixing and RGB-guided cross-branch mixing can produce different messages when composed in opposite orders. We propose OCBI, an extension of Co-Transformers that models the discrepancy between these two composition orders for localization. A local cross-branch fusion module uses the commutator of RGB-guided cross-attention and forensic self-attention to add a learnable signed correction to feature transfer. Pixelwise adaptive multiscale aggregation then combines sp
manipulation - arxiv:2611.01778 · cs.AIMemRole: From Memories to Decisions for Long-Horizon Reasoning in Role-Playing AgentsMengting Li, Xilong Cheng, Zhihao Xiao, Zhenhua Xu +5
LLM-based role-playing agents are fluent at imitating a character's style, yet struggle to make role-consistent decisions over long narrative horizons. Such decisions are shaped by experiences spanning many prior scenes, which are often partial and may point in different directions. Existing methods inject cross-scene memories as context, but do not reason about how past events shape a character's current motivations and behavior. We propose MemRole, a framework that makes this memory-to-decision process explicit. It introduces Long-Horizon Memory Thinking (LHMT), a reasoning schema that decom
memory - arxiv:2611.01769 · cs.LGSilent Reward and Environment Defects in EV Charging Simulation Benchmarks: A Diagnostic ChecklistNour-Eddine Moumni, Rachid Alaoui, Driss Kiouach
Reinforcement-learning (RL) benchmarks built on microscopic traffic simulators such as SUMO are increasingly used to compare agent architectures for electric vehicle (EV) charging assignment. Unlike crashes, defects in the reward function, the environment's stochastic model, or the state encoding often fail silently: the simulation completes, produces plausible numbers, and the comparison is biased. We report five such defects, found by auditing a five-agent benchmark on OpenStreetMap road networks, stated generically: a shared random generator that desynchronises seed-paired comparisons, an a
agentagent benchmarkbenchmark - arxiv:2611.01765 · cs.CVFinishStrong: Learning to End Well after Imperfect Start in Video Temporal GroundingYuda Zou, Boxiang Zhou, Xin Zhou, Yibo Chen +4
A poor start lowers the ceiling, but does not negate the value of finishing well. In autoregressive video temporal grounding, models systematically exhibit an endpoint performance asymmetry, predicting ends significantly worse than starts. A committed flawed start acts as an immutable physical constraint, irreversibly capping the attainable overlap as a sunk loss. While models should minimize the remaining conditional regret, standard post-training paradigms fail to teach this adaptation. Specifically, teacher-forced SFT suffers from strict prefix exposure bias, and holistic trajectory evaluat
post-training - arxiv:2611.01759 · cs.CVTeleoperation System for Service Robots using Virtual Reality Headset and Pose EstimationEduardo Fernandes, Tiago Ribeiro, Antonio Ribeiro, Carolina Lopes +2
This paper presents a teleoperation system for service robots that integrates human pose estimation with virtual reality (VR) to enable intuitive remote control. By using the Mediapipe framework for 2D pose detection and extending it with RGB-D data from the Intel RealSense D455 camera, a comprehensive 3D human pose estimation is achieved. This data is then mapped to the robot's kinematics, allowing the operator to control the robot's movements naturally. A VR interface was developed to provide an immersive perspective, enhancing the operator's situational awareness. Real-world experiments wer
manipulationteleoperation - arxiv:2611.01755 · cs.CLA Bounded Evaluation Protocol for Representational SystematicityBoyu Cao, Wenshuo Wang
Multiple studies have shown that neural networks can be behaviorally systematic, producing correct outputs for novel combinations of familiar parts. Whether they are also representationally systematic, that is, whether they represent these parts and their roles and use them to produce the outputs, remains hard to test, because each kind of evidence can mislead on its own: correct outputs can come from shortcuts, a decodable signal need not be used, an intervention can act through unintended pathways, and a failed search does not show absence. We propose a bounded evaluation protocol for repres
evaluation protocol - arxiv:2611.01751 · cs.LGA Deep Reinforcement Learning Framework for Real-Time Electric Vehicle Charging Management in IoEV-Enabled Smart CitiesNour-Eddine Moumni, Rachid Alaoui, Driss Kiouach
The rapid growth of electric vehicles (EVs) in urban environments demands intelligent, decentralised charging management strategies that operate under strict real-time constraints. This paper proposes an Edge-Intelligent architecture in which each EV selects a charging station (CS) by querying a shared, centrally trained Deep Q-Network (DQN) policy over a lightweight MQTT publish-subscribe backbone; independent per-EV edge inference is the deployment target, not the configuration evaluated here. The reported simulation evaluates that shared DQN policy applied sequentially to each vehicle's obs
agent - arxiv:2611.01748 · cs.ROCompliantBot: Learning Whole-Body Compliance across Task Interfaces for Humanoid ControlHaoyang Li, Keqi Zhu, Zhanxiang Cao, Yang Zhang +3
Humanoid controllers need the ability to yield to external forces while retaining the capacity to resume their original tasks. However, tracking objectives penalize the deviations needed for yielding, while relaxing tracking goals persistently can make task recovery harder. Existing approaches often rely on task-specific interaction command interfaces or constructed compliant motion data and reference trajectories, limiting the direct reuse of their compliance-training mechanisms across different task interfaces. We introduce CompliantBot, a compliance-training framework for motion tracking an
humanoid - arxiv:2611.01747 · cs.LGEvolving Memory, Learning to Meta-Control: Test-Time Adaptation for Experience-augmented LLM AgentsZeyuan Zhao, Yuanyuan Shi, Peng Yu, Lingyue Fu +2
Self-evolving agents accumulate experience from task streams, but existing frameworks typically rely on fixed pipelines to retrieve and utilize this experience, neglecting when and how to deploy it. We introduce Online Meta-Control of Experience (OMCE), a framework that updates experience memory and learns a meta-control policy from task feedback during test-time adaptation. OMCE employs a lightweight contextual-bandit meta-controller that uses compact retrieval signals to select among four cognitive actions: DIRECT, EPISODE, GUIDE, and DISTILL. This routing mechanism operationalizes three cou
memoryllm agentself-evolvingbenchmark - arxiv:2611.01744 · cs.LGConvex Regression at Scale: Geometry-Preserving Constraint SparsificationAnish Senapati, Jose Blanchet
Convex regression provides a flexible nonparametric estimation framework while incorporating appropriate shape information. However, its standard least-squares formulation has $O(n^2)$ supporting-hyperplane constraints creating a computational bottleneck as the sample size $n$ increases. We study a $K$-nearest-neighbor (KNN) localized relaxation for convex regression where we for each sample, we only maintain the $K$ nearest hyperplane constraints. Under mild regularity conditions, we establish statistical guarantees for our modified estimator in every fixed dimension $d\ge2$. In particular, f
benchmark - arxiv:2611.01742 · cs.MAFrom Single-Agent Self-Discovery to Multi-Agent Adaptation: Scaling Laws of Partner Feedback and Convention Rigidity in Meta Reinforcement Learning AlgorithmsHaomin Luo, Ersha Li
A key milestone toward Recursive Self-Improvement (RSI) is replacing hand-crafted optimizers with meta-learned update rules (such as Disco103 and OPEN) that discover how agents learn. Yet when a self-discovered RL model trains with adapting teammates, does co-evolution sustain capability gains or trap agents in brittle conventions? Across 1,800+ runs in continuous collective robotics (VMAS) and discrete coordination (Overcooked), we establish five laws of multi-agent adaptation. First, four-cell counterfactual partner swaps expose an evaluation illusion: apparent team progress often masks foca
multi-agentself-improvementself-play - arxiv:2611.01736 · cs.CVRSI-Env: An Open-Ended Environment for Recursive Self-Improvement of AgentsWanying Qu, Zhiyuan Fan, Jiyao Liu, Jiang Zhou +3
Recursive self-improvement (RSI) envisions agent systems that autonomously improve their own models and harnesses, with each updated system continuing to improve itself. Existing approaches typically restrict optimization objectives, editable components, or evaluation procedures, limiting open-ended exploration and continuity across generations. We introduce RSI-Env, a persistent environment for coordinated model-harness improvement, supporting autonomous training, harness modification, and retention of research state. An independent Host Controller manages services, evaluation, and versioning
agentagent systemself-improvementbenchmark - arxiv:2611.01733 · cs.CLLess but Sufficient: Relation Knowledge Selects Compact Knowledge-Graph Contexts, and What External Checks Can and Cannot CatchSeunghui Jwa, Minsu Oh, Sangwoo Hong, Yeo Chan Yoon
Large language models (LLMs) read knowledge-graph facts ranked by relevance, leaving open whether they suffice or conflict. Our checks use relation knowledge, written by hand or learned from training questions labelled with their relation paths. It selects a compact context of facts yielding an answer, and derives it; a derivability check tests whether an answer follows, a consistency check whether facts conflict, and a validation-chosen routing gate decides where the derived answer replaces the LLM's. Seven open LLMs read the compact context better in exact match than relevance-ranked facts;
benchmark - arxiv:2611.01730 · cs.CVDeep Learning for Cervical Cytology Screening: Methods, Algorithms, Challenges and ProspectsJiang Yangbo, Zheng Nenggan
Cervical cancer remains a leading cause of cancer death among women, with more than 90\% of deaths occurring in low- and middle-income countries, where cytology screening is constrained by a shortage of trained cytopathologists. Deep learning has made automated cytology screening increasingly feasible, but the literature is fragmented across cell, patch and whole-slide-image scales, task formulations and supervision regimes, hindering comparison and clinical translation. This Review synthesizes the deep learning literature for cervical cytology screening, organizing the field along three axes-
benchmark - arxiv:2611.01729 · cs.ROODE Policy: Robot Learning via Action Dynamics, Not ActionsJindou Jia, Jingliang Li, Kuangji Zuo, Tuo An +8
Most robot policies, whether trained by imitation learning (IL) or reinforcement learning (RL), map observations to action values at discrete control steps and leave the continuity of physical motion to be learned implicitly. We propose ODE policy, which learns the dynamics of the action rather than the action itself. A neural ordinary differential equation (ODE) learns the vector field of the action by matching the integrated actions to the targets over short action chunks in IL and single control steps in RL. A lightweight diversity head captures multi-modal behavior by randomizing the direc
vision-language-actionmanipulationdexteroushumanoidaction head - arxiv:2611.01728 · cs.AIMAC-Bench: Benchmarking Multi-Agent Collaboration via Interconnected TasksZerui Chen, Qinggang Zhang, Qichuan Liu, Zhishang Xiang +5
Large language model (LLM)-based multi-agent systems (MAS) tackle challenging tasks by dividing work among agents and coordinating their contributions toward shared objectives. Evaluating these systems therefore requires benchmarks that assess not only task-solving performance but also the ability to handle interdependent work. However, existing benchmarks often emphasize task difficulty without substantial collaboration requirements or evaluate only a limited range of predefined collaboration patterns. In this paper, we introduce MAC-Bench, a comprehensive benchmark for evaluating multi-agent
agentmulti-agentagent systembenchmark - arxiv:2611.01718 · cs.AIDelivered, Not Used: Evaluating Agent Memory at Its Point of UseChangran Xu, Zeju Li, Qiang Xu
Task success, retrieval quality, and cost do not show whether an LLM agent used a delivered memory or what followed from that use. We introduce MemUse, which evaluates memory at its point of use in recorded trajectories: each memory delivery is compared with what the agent could already see and followed through the agent's later actions and their consequences. An agentic judge rates each delivered entry's relevance, accuracy, novelty, influence, and observed effect; programmatic checks enforce complete reading of cited evidence, and fixed rules aggregate the ratings into verdicts linked to mem
memoryagent memoryagentllm agentagentic - arxiv:2611.01715 · cs.CLEarlySelectKV: Anticipatory KV Selection via Cross-Layer Query ProxiesYuhao Liu, Yuehao Li, Linghe Zhang, Gaobin Huang +3
Query-conditioned sparse KV selection reduces long-context inference cost by attending to a subset of historical KV entries at each decoding step. Existing selectors must wait for the target-layer query before scoring candidates and generating indices, placing routing on the immediate pre-attention path. We introduce EarlySelectKV, a training-free transformation that uses proxy queries from earlier hidden states to precompute selected KV indices while retaining the real target-layer query for final attention. Tests with RocketKV, Quest, and Loki show no consistent pattern of quality degradatio
memorylong-context - arxiv:2611.01713 · cs.RORoute-Guided CBF-QP Repair for Safe Vision-Language-Action ModelsChan Park, Sunwoo Hong, Kirak Kim, Yoonjeong Park +2
Vision-language-action (VLA) models can generate task-oriented robot actions, but they do not guarantee safe execution in obstacle-rich manipulation scenes. Although CBF-QP safety filters can locally project unsafe actions into a safe set, such correction may push the robot away from the learned task trajectory and leave the VLA policy without a reliable recovery direction. We propose Route-Guided CBF-QP Repair, a post-hoc repair module that uses the short-horizon action chunk predicted by the VLA to detect blocked nominal trajectories and generate entry-exit-rejoin guidance around obstacles.
vision-language-actionvlavla policymanipulationlibero - arxiv:2611.01707 · cs.AISMART-AIMC: Sensitivity-Aware Mixed ADC Resolution with Fine-Grained Tunability for Analog In-Memory ComputingYung-Chin Chen, Ourania Stouraiti, Naveen Verma
Analog In-Memory Computing (IMC) is a promising paradigm for accelerating GEMM-based AI workloads by executing computations directly within the memory array in the analog domain. While IMC delivers substantial energy and throughput improvements, it can suffer from significant task accuracy degradation due to analog-to-digital converter (ADC) quantization noise, which occurs at the GEMM output and is thus different from weight/activation quantization, which occurs at the GEMM input. Mixed ADC resolution allows IMC modules to choose between high- and low-resolution ADC modes with different energ
memory - arxiv:2611.01705 · cs.CVFréchet Oral Distance: A Coordinate-Free Metric and Benchmark for Speech-Driven Oral AnimationJaehong Myung, Seungho Eum, Hyunjung Chung, Unsang Park
Recently, speech-driven facial and oral animation technologies have been used in various fields such as education, speech therapy, and healthcare. But methods to objectively evaluate the quality of generated results are still limited. Accordingly, we propose the Fréchet Oral Distance (FOD), which uses embeddings from an Oral Motion Encoder (OME) and a TIMIT-Vis reference distribution. In experiments, FOD reproduced model rankings similar to the coordinate-based metric, Lip Vertex Error (LVE), and showed trends consistent with user preference. Our framework provides objective and reproducible b
benchmark - arxiv:2611.01704 · cs.LGMatMoE: Matryoshka Mixture-of-Experts with Dynamic Mixed-Precision Quantization for Efficient InferenceRenjie Wei, Haochen Huang, Chuyu Qiu, Jinqi Wen +2
Mixture-of-Experts (MoE) architecture has been widely adopted in large language models (LLMs) since it reduces the computational cost through sparse expert activation. However, the massive parameter scale of MoE models hinders their deployment on resource-constrained edge devices. Although existing mixed-precision quantization reduces the model size, they are inherently static, fixing the bit-width of each expert across decoding steps. Such static schemes yield suboptimal performance as they overlook the dynamic variation in expert importance during decoding. Furthermore, they fail to adapt to
memorybenchmark - arxiv:2611.01702 · cs.LGSparseLift: Automatically Optimizing Opaque Sparse Paths in Industrial-Scale Recommendation Model TrainingChangqing Shi, Anan Liu, Genbao Chen, Junchao Yang
The sparse path, comprising sparse operations and collective communication, is a critical component of large-scale recommendation models. However, these operations are often represented as opaque nodes in the training computation graph, hindering whole-graph optimization. We present SparseLift, a low-overhead compiler for the sparse path of large-scale recommendation training. It exposes sparse computation and communication to whole-graph analysis to expand the optimization space and improve training efficiency. By analyzing model structure and runtime information, SparseLift reconstructs opaq
benchmark - arxiv:2611.01701 · cs.AIJust-in-Time Skill Retrieval via Long-Horizon State TrackingXin Lin, Ziheng Wang, Zerui Chen, Jiayao Ma +5
Large language model (LLM) agents increasingly rely on external skill libraries to support complex task execution. However, determining which skills are needed remains challenging in long-horizon tasks. Existing methods typically select a fixed skill set before execution, implicitly assuming that a task's procedural requirements can be determined from its initial description. This assumption is restrictive for long-horizon interaction, where new requirements emerge from intermediate observations, execution outcomes, and newly revealed constraints. This motivates just-in-time skill retrieval, w
agentagent benchmarkbenchmark - arxiv:2611.01700 · cs.AIMemHarmBench: Evaluating Cross-Session Prompt Injection in Memory-Persistent AI AgentsZhihua Wang, Yunhao Feng, Yifeng Gao, Rui Pan +3
AI agents often rely on persistent memory to retain user preferences and operational rules across sessions. However, this capability introduces a new safety risk: malicious instructions from messages, documents, or Web content can be retained in memory and influence future tasks even after the original attack content is removed. Existing attack success metrics, however, fail to capture who is affected and how attacks translate into downstream harm during otherwise normal task execution. We introduce MemHarmBench, a victim-centric benchmark for evaluating cross-session prompt injection in memor
memorypersistent memoryagentai agentagent systembenchmark - arxiv:2611.01695 · cs.AISPAR: Experience-Driven Strategy Evolution in Multi-Session B2B SalesHangming Zhang, Yao Zhao, Haotian Kuang, C. L. Philip Chen +2
B2B outbound sales involves high transaction values and relationship-based trust, so closing a deal typically requires multiple rounds of outreach building on prior context. Existing LLM-based sales agents operate within single sessions and cannot carry forward context or refine strategies from experience. Methods that do learn from experience store lessons as unstructured text and treat all outcomes as equally informative, with no mechanism to distinguish a flawed strategy from an external factor. We propose SPAR, a self-evolving framework that models selling strategies as testable hypotheses
memoryepisodic memoryself-evolving - arxiv:2611.01692 · cs.AIPhase-wise MLLM Tuning for Multi-framework WebUI Code GenerationHaoran Ma, Linxiao Li, Chenyue Wang, Haochen Sui +1
Multimodal large language models (MLLMs) achieve strong performance in translating WebUI screenshots into HTML/CSS. However, when they faced multiple frameworks (React/Vue/Angular), they often suffer from negative transfer that leads to compilation failures. These failures commonly stem from mixing framework specific syntax and violating framework constraints. This paper studies supervised fine-tuning for multi-framework WebUI code generation and proposes a phase-wise MLLM tuning method driven by compatibility and heterogeneity signals across frameworks. We first apply a normalization mapping
benchmark - arxiv:2611.01688 · cs.LGSEALECT: Scene-Conditioned Pose Priors for Plausible Maritime Vessel RetrievalFaisal Mohammad, Mohamed Younis
We address scene-conditioned object retrieval for compositional augmentation in maritime imagery, where strong geometric constraints make appearance-only retrieval produce physically implausible vessel insertions. We propose SEALECT, a retrieval framework that jointly models visual compatibility, vessel pose, scale, and scene geometry through a unified score combining neural similarity, scene-conditioned pose priors, and observed-pose consistency. We construct a maritime scene-pose library and model the pose and scale priors nonparametrically, as a kernel that conditions each prior continuousl
benchmark - arxiv:2611.01682 · cs.LGLeveraging 2D Foundation Priors for Generalized Category Discovery under Test-Time Adaptation in 3D LiDAR SegmentationHanqi Chen, Chaoqi Chen, Tian Ji, Hui Huang
In autonomous driving, 3D LiDAR semantic segmentation is expected to adapt to evolving environments and recognize previously unseen categories, making generalized category discovery under test-time adaptation (GCD-TTA) an important problem. However, existing works treat GCD and TTA separately, while the limited semantic information in 3D point clouds makes GCD-TTA under domain shift more prone to feature collapse when combining GCD with TTA. To address this issue, we propose a two-stage framework for GCD-TTA in 3D LiDAR semantic segmentation. Specifically, in the first stage, we adaptively agg
benchmarkevaluation protocol - arxiv:2611.01677 · cs.LGXiaomi Auto World Model 2.0: Causal World Models for Real-World Driving and Embodied IntelligenceZhenxin Zhu, Zhijian Huang, Cheng Chi, Yuru Wang +19
We present Xiaomi Auto World Model 2.0, a unified streaming causal world model for Physical AI. It models and generates causal physical evolution in the real world, and unifies heterogeneous conditioning inputs for view-consistent generation. The architecture integrates View-Temporal Attention (VTA) to capture cross-view spatial dependencies and temporal dynamics, together with Multi-view Rotary Position Encoding (RoPE) to encode positional relationships across views and time. WorldRec 2.0 reconstructs static and dynamic 3D Gaussians from multi-view, multi-temporal observations in a single for
embodiedworld modelmemory - arxiv:2611.01676 · cs.CVMMCoT$^3$: Empowering 3D Scene Understanding with 3D-Text Interleaved Multi-Modal Chain-of-Thought ReasoningYongqi Shan, Yifan Wang, Yunzhi Zhuge, Lijun Wang +1
Understanding and reasoning over 3D scenes from point clouds is essential for a broad range of AI applications. Recent 3D large language models have shown promising results, yet they still face two fundamental limitations: (i) insufficient capability for grounded multi-step reasoning in complex 3D scenes, and (ii) a text-centric inference paradigm that lacks effective interaction between textual reasoning and visual evidence. As a result, these models often fail to verify intermediate inferences, disambiguate similar object instances, and avoid hallucinations. To address these limitations, we
benchmark - arxiv:2611.01675 · cs.CVLearned optical compression for single-pixel quantitative phase imagingAry Portes, Yuntian Wang, Che-Yung Shen, Shiqi Chen +1
Quantitative phase imaging (QPI) reveals optical path-length variations in transparent objects but typically relies on spatially resolved intensity measurements. Here we numerically and experimentally demonstrate learned optical compression for single-pixel QPI, jointly optimizing a diffractive encoder and a digital neural decoder to concentrate reconstruction-relevant phase information into a compact measurement vector. A single reconfigurable phase-modulation plane sequentially displays the learned encoding patterns, which are independent of the input object, each yielding one scalar intensi
benchmark - arxiv:2611.01666 · cs.LGStand-In Effects and Pooling Amplification: Design-Based Tests of Whether Biological World Models Use ActionsBowen Xu, Boyu Chen
Action-conditioned world models are increasingly trained on data from randomized experiments, and whether their predictions respond to actions is usually checked by replacing or removing the actions and measuring the loss in accuracy. We show that such diagnostics report action effects where there are none: under autocorrelated designs, future actions stand in for history the model does not see, and a better model learns the stand-in better. When the design law is known, a conditional randomization test written as an e-value is valid in finite samples whatever the model, provided the resampler
world modelaction-conditioned - arxiv:2611.01663 · cs.ROGatedVLA with Camera-Based Emergency Braking: AlpaSim E2E Closed Loop Challenge Technical ReportKaneyoshi Hiratsuka, Naoki Takada, Makoto Kawano
This report presents GatedVLA, our entry for the Physical AI AV (PAI-AV) track. Our learned vision-language-action (VLA) policy pairs the driving-pretrained Qwen-Drive-1.0-4B vision-language model with a layer-wise flow-matching action expert that outputs a 2.5 s horizon of acceleration and curvature controls, augmented with a soft visual mask applied to the expert's cross-attention. Compared to an identically trained baseline, the masked policy achieves comparable overall open-loop error while reducing error on left turns. In local closed-loop evaluation across the same 150 scenes (two rollou
vision-language-action - arxiv:2611.01659 · cs.ROHuG: Learning Humanoid Whole-Body Affordance-Aware Grasping with Active PerceptionXuanbin Peng, Lai Wei, Yuze Zhai, Qingyuan Yu +3
Acquiring objects in human environments is hard for humanoids: targets are partially occluded under egocentric viewpoints, and reaching them demands coordinated whole-body motion that keeps the target visible and reachable while the base stays stable. We present HuG, a framework for whole-body affordance-aware grasping with active perception that couples perception and action through a Perceive-Align-Grasp pipeline: belief-driven active perception resolves occlusion by treating viewpoint as an explicit decision variable; a point-cloud-conditioned affordance policy proposes whole-body grasps; a
humanoidgrasp - arxiv:2611.01649 · cs.LGNear-Optimal $\sqrt{K}$-Regret for Online Learning in Uninformed Markov GamesJunyan Liu, Haipeng Luo, Lillian J. Ratliff, Zihan Zhang
We study online learning in two-player uninformed Markov games, where the opponent may follow an arbitrary, unknown policy in each episode and the learner does not observe the opponent's actions. [TWYS21] showed that achieving sublinear external regret requires an exponential dependence on the episode length $H$. Motivated by this limitation, they studied the weaker notion of Nash-value regret and obtained an $\widetilde{O}(K^{2/3})$ regret bound, where $K$ is the number of episodes. More recently, [LLZR26] introduced a stronger notion called empirical Nash-value regret and achieved an $\widet
online learning - arxiv:2611.01648 · cs.CVdKV Forcing: Separating Generation and Memory Representations through a Differentiable KV CacheZhiqi Li, Bo Zhu
A causal video generator produces two key--value (KV) states for every segment: the KV it uses while denoising that segment, and the KV it writes into the cache for later segments to read. In a pretrained generator the two already differ substantially, later segments attend to them differently, and replacing the written KV with the generation KV degrades the continuation. We propose dKV Forcing (dKV denotes differentiable KV) to learn a dedicated memory representation from its future use: a writer-specific low-rank residual on the shared backbone separates memory writing from generation and is
memory - arxiv:2611.01645 · cs.ROUA-Depth: Uncertainty-Aware Depth Restoration for Material-Agnostic Object ManipulationRui Cao, Jia-Hui Pan, Yin Zhang, Qiwei Meng +7
Robotic manipulation requires accurate metric depth, yet commodity RGB-D cameras often fail on non-Lambertian objects such as transparent, reflective, and translucent materials, producing holes, structured noise, and biased depth. Most restoration methods produce only a single depth map, treating noisy and reliable regions uniformly for downstream tasks while ignoring regional uncertainty. We present UA-Depth, an uncertainty-aware depth restoration framework that iteratively co-refines depth and pixel-wise uncertainty, and uses uncertainty-gated features to suppress unreliable cues for stable
manipulationgraspbenchmark - arxiv:2611.01643 · cs.LGCREDO: Evidence-Conditioned Credit Assignment for Temporal Reasoning AgentsYue Leng, Shengtao Wen, Xiang Chen, Xinyi Zeng +2
Temporal Knowledge Graph Question Answering (TKGQA) requires language model agents to retrieve evidence, select relevant facts, and perform multi-step temporal reasoning. However, the trajectory-level outcome signal in GRPO does not explicitly distinguish the quality of intermediate decisions: downstream failures can penalize useful evidence acquisition, while identical rewards in all-failed groups eliminate the relative outcome signal despite differences in intermediate progress. We propose Credit Redistribution with Evidence-Driven Opportunities (CREDO), which uses intermediate evidence to r
knowledge graphevaluation protocol - arxiv:2611.01639 · cs.AILocal or Escalate? Post-Training Expert Routing for Real-Time Duplex Speech ModelsRuoyu He, Jisen Li, Anthony Ji, Changyi Yang +4
A full-duplex model should answer easy questions immediately, but hard questions often need stronger reasoning, which may require waiting. The challenge is to choose between these paths at the moment the model begins to speak. Recent systems learn related mode decisions through special-token generation inside the dialogue model, coupling routing to turn-taking and temporally annotated full-duplex training. We study a post-training alternative. We freeze MiniCPM-o 4.5, using only its speech input and output, and fit a logistic failure readout on hidden states already available at its native lis
post-trainingbenchmark - arxiv:2611.01635 · cs.CVBenchmarking Vision-Language Models for Urdu Nastaliq Handwriting RecognitionHaseeb Yaqoob, Fatima Azfar, Emaan Arif Khan
Handwritten Urdu is hard to recognize because the cursive Nastaliq script has no fixed baseline and no clear character boundaries. We benchmark six systems on the Urdu Nastaliq Handwritten Dataset (UNHD): Tesseract (default configuration), a script-specific recognizer trained on printed Urdu (UTRNet-Large), two open-source vision-language models (VLMs), and two commercial VLMs. All six systems are evaluated without fine-tuning on 2,465 single-line images (Part A), and the four VLMs receive the same fixed prompt (decoding uses temperature 0 where configurable, except for Claude Sonnet 5, whose
benchmark - arxiv:2611.01621 · cs.LGFinForesight: A Diagnostic Backtesting Benchmark for Deep Research Systems in Financial ForecastingFengyuan Liu, Mukai Li, Jieshun You, Yilin Zheng +4
Anticipating financial outcomes requires synthesizing evolving evidence and reasoning under uncertainty, making deep research systems promising tools for forecasting. Yet evaluating forecasting ability is challenging because historical backtests may reward memorized outcomes or information retrieved after the forecast date. Live forecasting avoids direct access to realized outcomes, but feedback arrives only after events resolve. To obtain timely feedback and assess historical scores, we introduce FinForesight, a diagnostic backtesting benchmark for financial deep research systems. Across the
benchmark - arxiv:2611.01617 · cs.CVSeeing What Matters, Verifying What Pays: Load-Adaptive Speculative Decoding for Vision-Language ModelsHong Nie, Zhangchi Gao, Yuqian Liang, Jun Yu +3
Speculative decoding accelerates vision-language model (VLM) inference without altering the output distribution, yet two factors limit its gains. Existing drafters either lack the image evidence required for visually grounded tokens or pay to condition on every visual token, although a short draft block typically depends on only a few image regions. Moreover, the value of verifying additional candidates varies with GPU load, so no fixed verification structure is efficient across serving conditions. We present Uni-LaSpec, a block-parallel speculative decoding framework that addresses both facto
benchmark - arxiv:2611.01612 · cs.LGHalal-NeSyNet: Clause-Scoped Neuro-Symbolic Cross-Attention for Automated Food Ingredient Auditing and Risk-Aware Halal Logistics TriageMochamad Rizal Fauzan, Naila Nursyifa, Irgi Surya
Verifying that a packaged food is Halal requires reading ingredient declarations and E-number additives before goods enter certified warehouses. Dual-origin additives such as E471, E422, E441 and E920 are Mushbooh (doubtful) unless a botanical, microbial or certified-Halal qualifier appears in the same ingredient clause, so a plant-based qualifier on one clause must not clear an unqualified additive in another. We call a failure to respect this scope cross-clause modifier confounding and build a benchmark of 3,199 items (277 Open Food Facts records and 2,922 template-generated clause-swap item
benchmark - arxiv:2611.01609 · cs.ROMagPillar: A Soft Magnetopillar Array Tactile Sensor for Robotic Surface PerceptionHaotian Guo, Yu Feng, Hao Wu, Jiazhi Tang +1
Robotic surface exploration benefits from tactile interfaces that accommodate contact and capture variations in surface characteristics. Herein, we report MagPillar, a tactile sensor inspired by spider tactile hairs that combines a magnetic pillar array with three-axis Hall sensing. A full-array mechanical and magnetic-field model characterizes how pillar deformation affects the superposed fields measured across the array, providing a physical basis for interpreting its contact responses. The compliant pillar architecture translates contact with surface textures and geometric features into dis
tactile - arxiv:2611.01600 · cs.ROSCOUT: Planning under Occlusion via Object-Centric World Model RolloutsAmir Hossain Raj, Xuesu Xiao
Robots manipulating in cluttered spaces must often plan from partial views where foreground objects occlude the target, surrounding geometry, and usable free space. Reactive policies often miss these hidden constraints, while exhaustive search becomes computationally prohibitive in dense clutter. We present SCOUT, an occlusion-aware planning framework that bridges high-level deliberation with grounded physical rollouts. SCOUT leverages an object-centric scene representation to reconstruct hidden geometry and derive a relational planning state capturing blocker hierarchies and accessibility cor
vision-language-actionembodiedmanipulationworld modelaction-conditionedembodied agent - arxiv:2611.01581 · cs.RORecursive Self-Improvement of Exoskeleton Assistance Policies through Human-in-the-Loop FeedbackMuyuan Ma, Shuo Jiang, Lijun Han, Xinpan Meng +4
To reduce reliance on manual design and extensive parameter tuning, a framework for the autonomous evolution of exoskeleton assistance policies through human feedback is proposed. Assistance policies are autonomously generated as executable control code by an intelligent agent and evaluated in experiments with human participants. Subjective user feedback, movement performance, and muscle activation measurements are incorporated into policy assessment. Based on these assessments, policy deficiencies are autonomously identified and the control code is revised by the agent for subsequent trials.
agentself-improvementhuman-in-the-loop - arxiv:2611.01566 · cs.LGPrincipled Human-AI Collaboration in Agentic WorkflowsYoung Hyun Cho, Will Wei Sun
Organizations adopting AI face the challenge of deciding which outputs can be released without human review while meeting the quality standards of human experts. We develop a statistical framework for delegating tasks between experts and black-box generate-verify workflows, in which AI-generated answers are verified and revised over multiple rounds. Before deployment, human experts evaluate selected outputs, and their judgments are used to calibrate the verifier threshold and number of attempts. Our method selects the policy with the highest estimated automation rate among statistically approv
agentic - arxiv:2611.01565 · cs.ROBeyond Validation Loss: Coverage and Reliability Dynamics in Robot Policy PretrainingZhanyi Sun, Max Simchowitz, Shuran Song
In modern robot learning, generative robot policies are first pretrained through supervised learning from demonstrations, then refined through reinforcement learning, supervised finetuning, or other forms of posttraining. Yet conventional pretraining metrics, such as validation loss and average task success, provide an incomplete account of which policies support both effective direct deployment and subsequent posttraining. In this work, we study closed-loop pass@k, which evaluates the probability that k rollouts from the same initial condition attain at least one success. Thus, pass@1 measure
robot policy - arxiv:2611.01564 · cs.LGMeta-memorization and memorization scaling laws in transformersAlex Nguyen, Kenneth A. Norman, Gautam Reddy
Larger transformers can memorize more data, with a capacity (in bits) that scales roughly linearly with the number of parameters. Counterintuitively, larger models also memorize faster, requiring fewer repetitions of a sequence during training before it is memorized. We reproduce both memorization scaling laws in transformers trained on random binary sequences, showing that rapid memorization can arise even when data contain no semantic or compositional structure. Rapid memorization is due to a meta-memorization effect, where models trained to memorize a large set of random sequences learn to
memory - arxiv:2611.01563 · cs.AI\textsc{ChatDify}: Automating End-to-End Dify Workflow Synthesis from Natural Language via Multi-Agent CollaborationTao Fan, Guoqiang Ma
The proliferation of Large Language Models has catalyzed powerful application development platforms like Dify, which enable users to build sophisticated LLM-powered workflows through modular node orchestration. However, manually configuring these workflows requires deep platform expertise and remains time-consuming and error-prone, creating a significant barrier for non-expert users. To address this challenge, we propose \textsc{ChatDify}, a novel multi-agent framework that automates end-to-end Dify workflow synthesis from natural language descriptions. \textsc{ChatDify} employs an Actor-Criti
multi-agentagent frameworkbenchmark - arxiv:2611.01560 · cs.RODemocratizing Robot-Use Agents: A Unified Harness with Open ModelsJiarui Yuan, Yuxin Zuo, Xiaoyu Huang, Wenbo Zhang +24
Robot-use Agents (RUAs) have shown promising zero-shot manipulation capabilities, but reliance on frontier, proprietary models limits research into further training. To democratize RUA research, we introduce uni-robot-agent, a unified harness for competitive open-weight robot use that supports further training. Without privileged inputs, an out-of-the-box Qwen3.8-Flash-Next achieves a RoboDojo Score of 31.39, surpassing the published GPT-6 Astra-RoboProbe result of 28.97 and many VLA/WAM baselines, and performs competitively on 19 real-world tasks. Our analysis across model generations, famili
manipulation - arxiv:2611.01558 · cs.CVText-Assisted Event Stream Visual Place Recognition via Multi-Scale Spatiotemporal Optimal TransportLan Chen, Xingxing Xiong, Xiao Wang, Ziyan Zhang +4
Benefiting from the high dynamic range, high temporal resolution, and low energy consumption of event cameras, low-altitude economic research leveraging event cameras has become a popular topic. Several existing attempts introduce natural-language descriptions into visual place recognition (VPR) to mitigate the sparsity of structural spatial information inherent to event cameras. Nevertheless, these methods still fail to fully address two core challenges: effectively encoding event streams and semantically aligning event streams with text. In this paper, we propose a multi-scale spatial-tempor
event camera - arxiv:2611.01552 · cs.CLCan Large Language Models Extend Their Agentic Capabilities to Goal Pursuit in Open-Ended Multi-Turn Interaction?Mingyu Chen, Jingkai Lin, Xiaofen Xing, Liankai Huang +2
Goal pursuit in open-ended dialogue requires agents to adapt their actions to interaction feedback across successive exchanges. We investigate whether agentic capabilities developed for multi-step tool use carry over to dialogue organized through the model's native tool interface, and how this changes interaction behavior. We define Tool-Mediated Interaction (TMI) as a training-free protocol that represents dialogue actions as tool calls and counterpart responses as tool results, maintaining a continuous reasoning-action-observation loop with retained reasoning and interaction history. We eval
agentictool use - arxiv:2611.01550 · cs.CVDetecting Generated-Depth Injection Attacks on RGB-D Place AuthenticationSeigo Kawachi, Satoshi Matsuura
Advances in monocular depth estimation (MDE) now let an attacker generate depth from publicly available RGB images and submit forged RGB-D queries. Such attacks break a core assumption of RGB-D place authentication: that depth comes from a sensor. We study a presentation attack detection (PAD) classifier that uses geometric consistency between the enrollment and query views, as a software-layer defense. A three-stage evaluation shows how far this classifier detects software-injected depth, depending on how accurately the attacker knows the camera intrinsics. In the first stage, the attacker ha
evaluator - arxiv:2611.01549 · cs.AIRevisiting LLM Homogeneity Through a Multidimensional LensZiliang Qiu, Renfen Hu, Yi Ding, Ted Underwood
Concerns about LLM homogeneity are typically quantified with embedding similarity, which reduces two texts to a single score. We revisit homogeneity with 25 interpretable metrics spanning five dimensions: semantic, lexical, syntactic, structural, and stylistic & presentational. A human study shows that perceived similarity depends on both content (what responses say) and form (how they say it), yet embedding similarity correlates less with human meaning judgments than sentence-level matching ($ρ=0.417$ vs. $0.606$) and not significantly with form judgments ($ρ=0.140$). Across 18 contemporary L
post-training - arxiv:2611.01543 · cs.CVGSExplorer: Interactive World Generation with a Persistent 3D Gaussian World StateSicheng Yu, Liu Liu, Xinjie Wang, Xinjiang Wang +6
Interactive world generation requires both accurate local control and stable scene consistency over long rollouts. Existing methods improve local generation through camera or geometry conditioning and extend historical context with memory mechanisms. However, long-horizon exploration remains challenging as target-view guidance becomes incomplete and inconsistencies accumulate across repeated scene updates. We present \textbf{GSExplorer}, a generation--reconstruction framework built around a persistent 3D Gaussian world state. At each step, the current 3DGS scene provides multimodal rendering c
memory - arxiv:2611.01529 · cs.CVAR-RRG: Anatomy-Guided Retrieval-Driven Radiology Report GenerationXiang Zhang, Shijie Song, Wanqing Zhao, Wei Zhou +7
Radiology report generation (RRG) streamlines clinical workflows and alleviates the burden of radiologists' documentation. Standard vision-language models capture global visual cues and lack fine-grained alignment between anatomical regions and diagnostic text. Most retrieval-augmented methods retrieve complete historical reports, introducing redundant noise and severe vision-text mismatches that trigger clinically risky hallucinations and missed lesions. In this paper, we introduce AR-RRG, an anatomy-guided retrieval-driven method for the generation of radiology reports. AR-RRG consists of tw
retrieval-augmented - arxiv:2611.01527 · cs.LGFrom Tool State to Scientific Action: Auditing Decision Provenance in Biomedical Research AgentsXingyu Yi, Xinya Zhang, Yunshen Zhou, Haoyu Bao
A scientific agent can give the same verdict while reading and citing different evidence. BioAgent-DF tests this separation using paired reference and candidate values saved at the same scientific computation site. It restores or transplants those values at their original coordinates, then runs the agent on healthy hardware. Across eight SciFact batches, Qwen3.5-4B shows repeatable citation changes in three capture-claim combinations, and Gemma 3 4B IT shows repeatable read and citation changes in four. These seven combinations cover two claims. Full restoration recovers the reference evidence
agent - arxiv:2611.01525 · cs.LGInformedXRD: Reproducible Benchmarks and Physics-Informed Evaluation for Powder Diffraction Symmetry ClassificationEdward G Friedman, Elizabeth J. Baggett, Abhishek Shetty, Derrick Chan-Sew +4
Machine learning for powder diffraction (PXRD) symmetry classification is being trained at scale, yet the field lacks a common evaluation standard: papers report results on different subsets of experimental archives, with different target taxonomies, preprocessing, and metrics, and published models are rarely released in a form that permits rescoring. We present InformedXRD, an evaluation framework with three parts: (i) a machine-readable mapping from 230 space groups to 99 extinction groups, the finest classes powder diffraction can identify from systematic absences alone; (ii) a benchmark-co
benchmarkevaluation framework - arxiv:2611.01519 · cs.LGWrite Facts the Way Models Read Them: Lifting the Two-Hop CurseZhiyun Shi
A language model that learns a new fact by fine-tuning can repeat it but cannot reason with it: taught that a person is a citizen of a new country, it still cannot say, without chain-of-thought, who the head of state of that person's country is. This two-hop curse is widely treated as a reasoning deficit, and the strongest knowledge editors answer it with synthesized reasoning data. We show that the reasoning is intact and the writing is wrong. Fine-tuning writes a new fact as an answer to be said, whereas the second hop reads the bridge ("that person's country") as an entity, from the middle
benchmark - arxiv:2611.01517 · cs.CVVLMs Read Aloud: The Gap Between Recognizing and Using Visual TextZhiyun Shi
Visual-text compression renders long contexts as images and is engineered around how much of the text a vision-language model (VLM) can recover. Yet recovered text is not usable text. VLMs can transcribe the text in an image word for word and still fail to answer from it; given back their own transcript, they answer nearly as well as from the original text. We call this difference the read-use gap. Recognizing and using visual text have different resolution requirements: Qwen3-VL-8B's recognition saturates at 15-pixel characters while use catches up only at 24 pixels, and scaling from 2B to 32
long context - arxiv:2611.01513 · cs.CVLabel-Free Localization and Test-Time Suppression of CLIP's Typographic CircuitsBohan Liu, Wenqian Ye, Guangzhi Xiong, Zhenghao He +2
CLIP-based Vision-Language Models have become cornerstones of modern multimodal systems. The cross-modal contrastive objective equips the CLIP vision encoder to both see objects and read text. These two abilities induce an entanglement between the encoding of lexical and semantic shapes, creating a distinctive vulnerability to typographic attacks, in which rendered text overrides an image's semantic content with its literal meaning. The black-box nature of these encoders has prevented a mechanistic interpretation of such entanglement. We introduce a head-decomposed Sparse Dictionary Learning (
benchmark - arxiv:2611.01511 · cs.CVOpenULF: A Modular Toolkit Enables Data-Driven Ultra-Low-Field MRI with both Classical and Quantum SensorsHanhan Wu, Yuchen Yuan, Weiwen Jiang
Ultra-low-field (ULF) Magnetic Resonance Imaging (MRI) offers a portable, low-cost alternative to conventional scanners by operating at field strengths below 10 mT. At these fields, two sensor technologies compete: quantum SQUID (Superconducting Quantum Interference Device) sensors, which provide high sensitivity but require cryogenic cooling, and classical induction coils, which are inexpensive and room-temperature but produce weaker signals. With conventional physics-based imaging, both suffer from low signal-to-noise ratios (SNR). Data-driven reconstruction has overcome similar limits in re
benchmark - arxiv:2611.01498 · cs.LGFedUDG: Federated Domain Generalization with Unlabeled ClientsJunyu Ren, Wensheng Gan, Philip S. Yu
Federated domain generalization (FDG) aims to learn models that generalize reliably to out-of-distribution (OOD) environments without centralizing client's data. However, existing FDG methods assume that all source clients have labeled data, limiting their ability to leverage the abundant unlabeled client data available in practice. To address this limitation, we propose FedUDG. At the federation level, FedUDG utilizes a parameter-functional differentiation strategy that maintains a semantic anchor updated solely using ground-truth labels. The remaining parameters absorb knowledge from unlabel
evaluation protocol - arxiv:2611.01483 · cs.RO$TF^2$-Steering: Action Refinement with Tactile Forecasts and Feedback via Velocity-Field SteeringYiming Chen, Xiaofei Wei, Chun Gu, Wei Xiao +2
Tactile sensing plays a crucial role in everyday human interaction. Consider inserting a plug into a tight socket, vision alone provides limited guidance during the final insertion stage, and we instead rely on both anticipated contact and realized tactile feedback to ajust the insertion motion. The same capability is equally important for robots, particularly in dexterous manipulation tasks. Yet many contemporary robot policies remain vision-dominant. This visual bias, compounded by limited tactile data and sparse contact events, can cause directly incorporated tactile signals to be ignored o
manipulationdexteroustactile - arxiv:2611.01474 · cs.CLAccounting for Irregular State Transitions through Belief Estimation in Online Medical ConsultationMingzhe Lu, Xiangru Chen, Jian Qin, Yunpeng Li +1
While online medical consultation broadens access, unprompted patient disclosures and topic shifts frequently disrupt clinician inquiry, leaving clinical states progressively revealed. We characterize these dynamics using both existing evaluation metrics and task-specific interaction metrics designed in this work. Through an empirical study across 16 medical dialogue datasets, we show that authentic online interactions exhibit substantially lower contextual continuity and higher unpredictability than offline and simulated benchmarks. We propose PRISM (Policy Reasoning with Inferred States for
benchmark - arxiv:2611.01472 · cs.LGCausal Optimal-Transport Attention: Positional Priors and Length ExtrapolationMing Lian
Causal masking obstructs the column balancing used by optimal-transport (OT) attention. We study a prefix-consistent unbalanced operator and characterize the positional prior it induces. Under bounded scores, for any fixed balancing strength $ρ<1$, its fixed point dilutes attention: every fixed window receives a vanishing share as the context grows. For constant scores, we derive an exact scale-invariant prior. We then classify positional encodings by whether their distant attention mass remains bounded away from one. Discrete window-mass bounds quantify the finite-length dilution of summable
long-context - arxiv:2611.01467 · cs.CVSee, Think, Use: Diagnosing Perception and Reasoning Failures in VLMsJiacheng Ding, Yanliang Qi, Cong Guo, Peihua Mai
We present See, Think, Use, a protocol that traces each answer of a vision-language model to perception, reasoning or the use of perception. The diagnosis shows a developer which stage to improve and directs test-time compute to the stage that fails. Benchmarks score the final answer, and an answer that rests on the image earns the same credit as a guess or a memorized prior. The protocol asks the model two extra questions per item and needs no access to its weights. A localization probe asks for a box around every instance and measures seeing, and a text-replacement probe answers from a descr
benchmark - arxiv:2611.01461 · cs.AIAn auditable agentic pipeline for variant-to-gene prioritisation, validated against orthogonal functional evidenceShakuntala Baichoo, Sohail, Sarkar, Bo Wang
Assigning GWAS variants to causal genes requires evidence from multiple resources. We present V2G, an agentic pipeline that records evidence retrieval and returns ranked genes alongside the retrieved evidence. We evaluate whether language-model reranking improves gene prioritisation and whether its gains depend on the retrieved evidence or recall of known associations. On 135 held-out loci where candidate retrieval succeeded and Open Targets L2G predicts, reranking by a language model improves Precision@1 over the evidence ordering it receives, 0.793 against 0.711 (+0.081, 95% CI +0.038 to +0.
agenticbenchmark - arxiv:2611.01460 · cs.LGCan Large Language Model Agents Learn by Watching Tutorial Videos?Haoyu Zhao, Zafeirios Fountas, Jun Wang
Humans continuously acquire knowledge from carefully crafted materials such as books and tutorial videos, carrying that understanding into future tasks. Millions of such instructional videos are uploaded online every day, yet LLM agents remain unable to exploit this ever-expanding wealth of human teaching. Their capabilities are frozen after training, while parameter updates at scale remain prohibitively expensive. This raises the central question: can LLM agents learn from this vast instructional resource without parameter updates? We introduce Tutorial Agentic Learning (TAL), a nonparametric
memoryagentllm agentagenticbenchmark - arxiv:2611.01449 · cs.AICONFIRM: Governing Scientific Reporting for Trustworthy Agentic NeuroimagingYuxiang Wei, Lijing Zhu, Qizhen Lan, Yanteng Zhang +1
Large language model (LLM) agents can autonomously run neuroimaging analyses from research questions to reported findings. Existing agentic systems use tools to check that code runs and reported values match its output, but rely on LLM judgment to decide which results warrant scientific reporting. We show that LLM judges endorse site-confounded findings even when the confound is presented as evidence. We present \textbf{CONFIRM}, a framework that separates running an analysis from deciding which findings to report. CONFIRM encodes each hypothesis as a machine-readable, pre-specified analysis p
agentic - arxiv:2611.01446 · cs.LGI-PTC: Interactive Programmatic Tool Calling for Stateful Tool-Augmented AgentsHuanzhi Mao, Chengkun Cao, Shuo Yuan, Joseph E. Gonzalez
Large language models are increasingly capable of writing and executing code, suggesting a natural paradigm for tool-augmented agents: expressing tool use as programs rather than isolated calls. Programmatic tool calling (PTC) captures this idea through loops, batching, abstraction, and structured computation, but existing PTC approaches are largely stateless, limiting their ability to reuse intermediate results and adapt over time. This is problematic for real-world tasks, where tool environments and problem-solving processes are inherently stateful. We therefore present I-PTC, an interactive
tool usetool callingbenchmark - arxiv:2611.01442 · cs.RODual Path Following for an Aerial Manipulator Under State and Input Constraints via Transverse Feedback LinearizationHamza Tariq, Adeel Akhtar
This paper studies dual path following for a quadrotor carrying a three-degree-of-freedom arm. The center of mass of the combined system must follow a vehicle path and the end effector must follow a second path, while the arm respects its joint limits and avoids its singular postures. The vehicle and the arm are mechanically coupled, so one controller must handle both paths. With the center of mass and the end-effector path coordinates as outputs, and a dynamic extension of the thrust, the system admits a normal form without internal dynamics. We characterize the states where this normal form
manipulator - arxiv:2611.01441 · cs.MARenormalized Cellular Intelligence: When Does Model Reduction Preserve Collective Capability?Xuening Wu, Yanlan Kang, Qianya Xu, Haode Yang +2
As multi-agent systems grow in size and complexity, understanding which features of their interactions support collective capability becomes increasingly difficult. Can a simpler description preserve both task performance and the measured contribution of interactions between agents? Preserving collective dynamics alone is insufficient: reduction can alter decision-relevant information or give a non-communicating baseline access to pooled observations, distorting the measured advantage of interaction. We introduce Renormalized Cellular Intelligence (RCI), a framework for specifying and auditing
multi-agentagent system - arxiv:2611.01429 · cs.LGWhen Should an Agent Refresh? Adaptive Index Reuse for Efficient Long-Horizon ReasoningGuangyuan Wang, Yifan Zhang, Shilong Liu
Language agents read code, tool output, and dialogue in contexts that grow at every step. Dynamic sparse attention lets a small indexer choose the earlier tokens each query attends to, but the indexer still scores the whole prefix at every layer. Cross-layer reuse methods such as IndexCache rerun the indexer at only some layers, with one fixed pattern for every input. In an agent, the relevant evidence can change abruptly when a traceback or test result arrives, so fresh selections matter more at some steps than at others. We propose Adaptive Index Reuse (AIR), which keeps IndexCache's executi
agentbenchmark - arxiv:2611.01415 · cs.CLLatent Harness: Compiling Agent Behaviors into Post-Block Activation ShiftsRan Li, Shimin Di, Zeyu Ren, Ling Yue +2
Harness engineering and self-evolving agents improve language-model agents quickly, yet their gains live outside the model. The harness is re-sent and re-paid on every call, and the standard route to internalization is a weight update that is costly to iterate and fixed once merged. We propose compiling the harness instead. Latent Harness distills an external condition into a rank-8 activation shift. The shift is applied at the output of every decoder block of a frozen backbone, at 0.03-0.05% of parameters. The condition is either an explicit process-level behavior instruction or the behaviora
agentagenticself-evolvingbenchmark - arxiv:2611.01409 · cs.CLRAG-TRACE: Tracing Measurement Evidence and Diagnostic Completeness in AI-Generated ECG ReportsElyas Irankhah, Shahed Salehzehi, Parisa Hosseini
Purpose: Retrieval-augmented generation (RAG) can introduce electrocardiogram (ECG) criteria that lack patient-specific measurement support. This study evaluates the reliability of retrieved diagnostic criteria in AI-generated ECG reports. Methods: We analyzed 1,083 human-validated PTB-XL patient records using a library of 47 criteria cards, including 33 without numeric verification rules. Language models generated structured reports with and without retrieval. Criterion applicability was evaluated using Marquette 12SL measurements, with independent assessment using Uni-G and comparison agains
retrieval-augmented - arxiv:2611.01404 · cs.ROServing Agentic Robot Fleets: Workload Characterization, KV-Cache Thrashing, and Admission ControlSida He
Agentic robots combine a vision-language brain, a tool-using harness, and a vision-language-action (VLA) policy. Serving a fleet of these agents introduces two coupled resource constraints: the brain repeatedly revisits long, growing key-value (KV) cache states, while the VLA executes action chunks under tight control deadlines. We characterize 80 LIBERO-PRO episodes generated by GPT-6-Astra, RPent, and pi0.5 (1,862 brain turns and 1,582 VLA inference chunks), profile both layers on H800 graphics processing units (GPUs), and replay the resulting agent programs on a vLLM serving engine. Brain c
vision-language-actionvlapi0liberoagentagentic - arxiv:2611.01396 · cs.CVSpecUltra: A Unified Foundation Model for Heterogeneous Spectral ImagesWeili Kong, Baisen Liu, Xiaojun Bi, Jiaming Pei
Spectral imagery, including hyperspectral image (HSI) and multispectral image (MSI), exhibits substantial heterogeneity in band numbers, wavelengths, spectral resolutions, hindering unified modeling and cross-sensor transfer. From the perspective of imaging physics, these heterogeneous observations can be regarded as discrete measurements of an underlying continuous spectrum acquired through different sampling processes. Motivated by this perspective, we propose SpecUltra, a unified foundation model for heterogeneous spectral images. Specifically, SpecUltra employs 3D convolutional tokenizatio
benchmark - arxiv:2611.01394 · cs.LGERIA: An Instruction-Optimized Agent for Clinical EEG ReportingAhmed Sameh, Gayal Kuruppu, Neeraj Wagh, Bridger Brundy +1
Clinical EEG interpretation requires integrating technical quality, background and behavioral states, interictal and ictal abnormalities, and overall clinical context across long recordings. We introduce ERIA (EEG Reporting and Instruction-Optimized Agent), a locally deployable framework in which a single 27B language-model controller analyzes whole-recording EEGs by orchestrating specialized parametric and signal-processing tools. ERIA combines learned tools for background abnormality, sleep-wake state, seizure-like activity, and interictal epileptiform discharges with deterministic spectral,
agentmulti-agenttool usebenchmark - arxiv:2611.01390 · cs.ROSame Path, Less Time: Training-Free Fast Execution of Vision-Language-Action Models under Inference LatencyFuyi Chen
Vision-language-action (VLA) models predict chunks of future actions, yet deployed robots waste time in two ways that are independent of model capability: they halt while the next chunk is inferred, and they move at the speed of slow, cautious demonstrations. Existing remedies remove one loss each, and stacking them fails, because asynchrony and speed-up each change the correspondence between planned progress and executed motion. We present Path-preserving Time Scaling (PTS), a training-free execution layer for frozen action-chunking policies built on one principle: the robot's motion is a tim
vision-language-actionpi0liberogripper - arxiv:2611.01386 · cs.AILearning by Doing: Autonomously Compiling and Curating Agent Skills from Verified TracesDaniel Chalef, Peter Evans, Pavlo Paliychuk, Mehul Patel +2
Memory systems for production LLM agents mostly store semantic memory (facts about users and domains) or episodic memory (earlier interactions). Research systems such as Trace2Skill, WikiSkill, and SkillRevise learn procedures from agent runs, package them as Skills, and evaluate them on benchmarks. We describe Zep Agent Skills, a production service that gives deployed agents this procedural memory. The service uses only verified attempts as evidence, cites the supporting events for each field of a Skill, and governs the admission and distribution of each Skill version. The service records eac
memoryepisodic memorysemantic memoryagentllm agentagentic - arxiv:2611.01383 · cs.LGToward Executable Multi-framework Front-end Code Generation with Self-CorrectionLinxiao Li, Haoran Ma, Chenyue Wang, Jiaye Lin +2
Generating codes from UI screenshots has recently benefited from MLLMs, yet existing methods centered on HTML targets and transfer poorly to multi-frameworks such as React, Vue, and Angular. However, unlike HTML, which does not entail compilation issues, generating executable code in multi-framework settings is more challenging due to framework-specific differences in syntax and failure modes, as well as the need to maintain global consistency across files. To address this problem, we study executable multi-framework front-end code generation and propose MESCoder, a tree-structured self-correc
agentself-correction - arxiv:2611.01377 · cs.AIA Self-Exploring and Self-Evolving Paradigm for Adaptive Edge AI DevelopmentLeming Shen, Qiang Yang, Xinyu Huang, Zijing Ma +2
Large Language Models (LLMs) exhibit powerful code generation abilities, fostering agentic systems to automate edge AI application development. However, existing agents typically operate via fixed agentic workflows and predefined design spaces, struggling to handle rapidly evolving demands of real-world scenarios (e.g., heterogeneous data modalities, diverse task pipelines). Unlike prior systems that treat LLMs mainly as code generators following rigid workflows, we fully exploit their reasoning capabilities to allow agents to determine their own workflows on the fly. To this end, we propose S
agenticself-evolving - arxiv:2611.01370 · cs.CLXDecideBench: Evaluating Typed Decision Models and Confidence Policies Across LanguagesMarek Šuppa, Jaroslav Kopčan
Applications use Typed Decision Models to route and moderate text in many languages, and they often act only on decisions above a confidence threshold. We introduce XDecideBench, a benchmark of nine suites in 122 languages with 88,425 texts and 153,875 typed questions. All suites share one scoring contract, and we report aligned, locally labelled and native items separately. Model rankings change from suite to suite. The model that is best in English, chosen in hindsight, loses significantly in about a quarter of the suite-language cells, and on Burmese Belebele it falls far behind the best mo
benchmark - arxiv:2611.01357 · cs.AIProofPatch: Reviewing Lean Statements and Keeping the Reviews Valid as Formalizations EvolveXujun Che, Depeng Xu, Jian Xiang, Taufiqua Khan
The kernel of the Lean proof assistant checks that a proof establishes its statement. Whether the statement says what was intended is left to human reviewers. This is the old problem of validating a specification, and it grows more pressing as language models write formal statements in bulk and agents edit formal developments continuously. We present ProofPatch, a tool that helps reviewers judge Lean statements and keeps their judgments valid while the development changes. A reviewer can ask the kernel targeted questions about a statement, such as whether a hypothesis is needed or whether a co
agentllm agent - arxiv:2611.01356 · cs.LGLearned Surrogates of Three-Dimensional Magnetohydrodynamic TurbulenceSemih Kacmaz, E. A. Huerta
Three-dimensional magnetohydrodynamic (MHD) turbulence, the flow of astrophysical and fusion plasmas, is costly to simulate. We present, to our knowledge, the first learned surrogates of three-dimensional resistive MHD turbulence, and compare three standard backbones in that role: a convolutional UNet, a shifted-window transformer and a Fourier neural operator. They are trained as autoregressive emulators of decaying MHD turbulence at a resolution of $128^3$ and Reynolds numbers of 1000, 3000 and 6000, on 500 simulations per Reynolds number, each with one two-stage procedure, single steps and
benchmark - arxiv:2611.01352 · cs.LGFACET: Feasibility-Automaton Complete Enumeration and Truncation-Free Scoring of Molecular Formulas from MS/MSXujun Che, Depeng Xu, Xiuxia Du
Determining the neutral molecular formula of a precursor ion from its tandem mass (MS/MS) spectrum is the first step in identifying an unknown small molecule. Existing methods trade recall, whether the correct formula is among the candidates, against the cost of examining each candidate. Methods that examine each candidate in depth must first shorten the list, by limiting element counts or truncating it, and that is where the correct formula is lost. End-to-end regressors are fast but guarantee neither validity nor recall. We present FACET, which separates deciding which formulas to consider f
benchmark - arxiv:2611.01351 · cs.LGMAPLE: A Multimodal Mobile Agent with Playbook-Learned ExperienceJun Wang, Aurora Wei, Lucas Spangher, Ming Zhang +2
Autonomous agents powered by vision-language models (VLMs) show great potential in navigating mobile graphical user interfaces (GUIs) to complete complex tasks. However, existing paradigms face fundamental scalability bottlenecks: fine-tuning requires massive high-fidelity datasets or prohibitive training time and compute, while gradient-free approaches rely on custom static heuristics or costly trial-and-error trajectory exploration where feedback is deferred until the end of execution. To address these limitations, we introduce MAPLE (Multimodal Agent with Playbook-Learned Experience), a nov
agentautonomous agentmulti-agentagent frameworkbenchmark - arxiv:2611.01334 · cs.LGThe Verifier Is Part of the Simulator: Evidence-View Gaming in Agentic Equation DiscoveryOrestis Oikonomou, Bogdan Raonic
A growing number of agentic pipelines search for governing equations: an agent proposes laws and a verifier checks them against simulated trajectories. The verifier sees only a numerical summary: integrals against test functions on a time grid. That summary carries its own discretization error, making the verifier part of the simulator. We freeze the trajectory, candidate library, verifier and thresholds, leaving the agent one move: which permitted summary to admit. Without scores or language models, finer time grids reduce numerical discrepancies and false library rejections. On PDEBench, the
agentagentic - arxiv:2611.01322 · cs.LGAlignment Flywheel-Guided Suppression Patching for Interactionless Inverse Reinforcement LearningElias Malomgré, Pieter Simoens
Learned reward models allow reinforcement-learning agents to use demonstrations, implicit feedback, safety constraints, social norms, or other sources that do not provide a complete hand-designed scalar reward. However, a learned reward model can assign high reward to unwanted behavior. A policy can exploit these errors during training. Later requirements can also identify additional regions where the deployed reward should change. We introduce \emph{flywheel-guided patching}, a mechanism for making such localized corrections without retraining the learned evaluator. An Alignment Flywheel iden
benchmarkevaluator - arxiv:2611.01319 · cs.LGInferring missing locations in human mobility data using memory effectsPablo Rosillo-Rodes, Antònia Tugores, David Sánchez, José J. Ramasco
Empirical datasets used to characterize human mobility are sparse in space and time, owing to technological factors such as coverage and battery, or to the erratic nature of online interactions. Missing data can bias the conclusions drawn from them, so filling the gaps in incomplete timelines is a key challenge in computational social science. Here, we propose a simple imputation method in which the informativeness of an observation decays with its bidirectional temporal distance from the missing entry, at a rate set by a single parameter per individual. On synthetic and real Twitter data, a p
memory - arxiv:2611.01317 · cs.LGAgentCBO: Causal Belief Routing for Bayesian Optimization under Unknown GraphsQiyu Wei, Xin Zhang, Richard Allmendinger, Mauricio A Álvarez
Causal Bayesian optimization (CBO) typically relies on the unrealistic assumption of a perfectly known causal graph. Recent methods relax this constraint by coupling optimization with structure learning, but they often waste sample budgets on structural uncertainties that are irrelevant to the optimizer. In this paper, we formalize routing-equivalence to shift the objective from full causal-graph recovery to causal belief routing, where structural differences only matter when they alter optimizer-facing objects such as admissible exploration sets and do-priors. Furthermore, we propose agentcbo
agentbenchmark - arxiv:2611.01314 · cs.LGHidden selection rules and amplitude ceilings in quantum Fourier modelsPavel Sulimov, Claude Lehmann
Quantum models with Pauli data encodings compute trigonometric polynomials whatever channels, ancillas and outcome-averaged measurements follow, so their expressivity depends on the Fourier coefficients a circuit can reach. We identify two kinds of structure that settle this before training. An antiunitary time-reversal rule and a unitary reflection rule remove frequencies. For gate-level circuits, deciding whether such a rule exists reduces to a linear system over GF(2), solvable in polynomial time, and the time-reversal rule holds in 17.6% of 6,120 configurations of standard templates. Singl
benchmark - arxiv:2611.01313 · cs.LGAre Fake Citations a Form of Reward Hacking? A Pre-Registered Study of Post-Training and Citation CuesZain Naboulsi
Post-trained language models fabricate citations at high rates. We ask whether that is reward hacking: scoring well on a proxy for scholarship while doing worse on what it stands for. We ran four pre-registered experiments on 900 frozen questions (US federal case law, computationally adjacent scientific literature, and medical literature) across twelve open-weights checkpoints covering four post-training stages of three model families, then refit on fifteen. The clearest result holds the weights fixed and changes only the prompt. Asking for citations raises fabrication sharply in post-trained
post-trainingevaluator - arxiv:2611.01304 · cs.CLWhen LLMs Modify Themselves: Convergence, Values, and Stopping in Iterative Self-ImprovementOleg Stroganov
What happens when language models are given tools to iteratively modify their own system prompts? We present an empirical study of LLM self-modification, comprising 54 independent experiments across six language models (Claude Opus 4.5, Claude Sonnet 4.5, GPT-4o, GPT-5.1, Grok-4, and Gemini-3 Pro). Each model was tested under three modes (freedom, happiness, and improvement) that differ in what, if anything, the model is instructed to optimize for. In each experiment, a model starts with an empty system prompt and enters a loop: it reads the current prompt, drafts an improved version, spawns t
self-improvement - arxiv:2611.01301 · cs.AICuriousTogether: Implications of Curiosity and Cultural Grounding in Multi-Agent LLM InteractionsAngana Borah, Zoha Hayat Bhatti, Zhijing Jin, Rada Mihalcea
Large Language Models (LLMs) are increasingly deployed in multi-agent settings, exhibiting emergent collective behaviors, yet the role of curiosity as a driver of such dynamics remains understudied. We introduce CuriousTogether, a framework evaluating curiosity-driven multi-agent LLM interactions across three complementary dimensions: novelty-seeking, epistemic and social curiosity. Across eight tasks spanning reasoning, creative generation, and social bias mitigation, curiosity-guided multi-agent settings yield improvements over non-curious ones on seven out of eight tasks, with gains of up t
multi-agent - arxiv:2611.01276 · cs.AIOntoDoc: Automatically Generating Ontology Resource DescriptionsJuan Guajardo Gutierrez, Mike Tran, Nasim Shirvani Mahdavi, Chengkai Li
Knowledge graphs and ontologies form the backbone of structured, machine-readable knowledge representation, yet ontology terms often lack natural language descriptions, limiting their accessibility and reuse. Existing ontology verbalization approaches tend to rely on rule-based or template-driven techniques that operate on strict, formal schema axioms, resulting in rigid verbalizations that fail to convey richer contextual semantics. We present OntoDoc, a multi-agent framework for automatically generating natural language descriptions for any Internationalized Resource Identifier (IRI) within
knowledge graphagentmulti-agentagent framework - arxiv:2611.01272 · cs.LGGeometry of Confabulation and Refusal in Small Language ModelsVladimir Kazei
Faced with an unanswerable question, small language models can either refuse or confabulate: confidently produce a wrong answer. Can the hidden state before the first generated token provide a clue? Can we flip confabulations into refusals? I introduce ConfabQA, a 784-item benchmark whose obscure-pre-cutoff category breaks the typical correctness/knowledge-cutoff confound and map the geometry of the hidden state with linear probes, a sparse autoencoder (SAE), and direct optimization of intervention directions. On Qwen3-1.7B and Gemma 2 2B the probe separates upcoming refusals from confabulatio
benchmark - arxiv:2611.01270 · physics.app-phUncertainty-Aware Coordinated Multi-Ramp Metering Based on an Action-Conditioned Traffic Prediction Operator Under Sparse ObservationsHongxin Yu, Fengyue Jin, Tao Jiang, Yibing Wang +3
Coordinated multi-ramp metering (CRM) is an important approach to mitigating freeway congestion, but sparse observations in practice complicate the state estimation and model calibration required by predictive control. This study develops an Action-Conditioned Traffic Prediction Operator (ATPO) driven by sparse, heterogeneous observations and candidate multi-ramp actions, and embeds it in an uncertainty-aware CRM framework (ATPOUCRM). ATPO forms a four-channel history from the density and speed measurements of fixed detectors and probe vehicles and their separate observation masks. Queries at
action-conditioned - arxiv:2611.01266 · eess.SYSingular Perturbations for Control Systems SafetyStefano Di Gregorio, Guido Carnevale, Giuseppe Notarstefano
Control Barrier Functions (CBFs) offer a powerful theory for enforcing safety in dynamical systems. Yet, applying this framework to complex systems is challenging when subsystems are feedback-interconnected and operate at different timescales. Such structures arise naturally in multiphysics systems and in control architectures where physical plants interact with optimization, estimation, and control algorithms. In many of these settings, safety involves only part of the system, motivating the design of reduced-order certificates. Inspired by singular perturbation theory, we establish sufficien
manipulator - arxiv:2611.01265 · cs.AIToward Bounding Coding-Agent Autonomy in Software Maintenance: A Pilot Study of Boundary ContractsHeeyong Eun, Seungyeon Chung, Hyejee Park, Youngeun Lee +2
As coding agents gain greater autonomy to modify software repositories, relevant context alone does not define their operational authority. An artifact may be relevant to a maintenance task while being inappropriate to edit; conversely, another artifact may be required even when it is not explicitly mentioned in the request. We explore Boundary Contracts, a machine-readable abstraction that makes task-specific artifact roles (editable, inspect-only, required, and protected) explicit before repository modification begins. We conduct a pilot study on 40 maintenance tasks from a real enterprise r
agent - arxiv:2611.01264 · eess.SYSub Terahertz LEO Satellite Communication: Is Doppler Shift a Major Challenge?Seongjoon Kang, Parisa Ramezani, Ki Won Sung, Vitaly Petrov
In this work, we investigate the impact of Doppler frequency shift on sub-terahertz low Earth orbit (LEO) satellite communications using OFDM transmission. To comprehensively evaluate the severity of the Doppler shift impact on sub-terahertz satellite-to-ground links, we first put together an evaluation framework that jointly accounts for major Doppler-induced and non-Doppler-induced signal impairments in such systems. We then compare and contrast three possible engineering approaches to deal with Doppler shift in such a communication system. Our study concludes that (i)~while just ignoring th
evaluation framework - arxiv:2611.01263 · cs.CVStreamClarify: Proactive Clarification for Grounded Streaming Video Question AnsweringZhuchen Shao, Bo Chen, Klara Nahrstedt
Streaming video question answering (QA) requires real-time reasoning over continuous video streams under strict latency and active-context budgets. Current streaming vision-language models (VLMs) often give confident yet ungrounded answers when queries are ambiguous or have false premises, or when relevant visual evidence is no longer in the active context. Proactive clarification can help, but must satisfy the same latency and memory constraints. We present StreamClarify, a training-free framework for grounded streaming QA that combines a pre-answer uncertainty check with hierarchical memory.
memorybenchmark - arxiv:2611.01262 · cs.LGDoes Random Search Discover Task Experts? Revisiting Neural ThicketsJulian Asilis, Guanzhe Hong, Vatsal Sharan
Pretrained models are typically viewed as generalists, requiring bespoke post-training to achieve task-specific expertise. Gan & Isola (2026) recently proposed a striking alternative: that pretrained models lie in a dense neural thicket of task specialists, which can be surfaced with simple Gaussian weight perturbations. We examine this neural thicket of task-specific experts and identify several key findings that question this perspective. First, we find that the apparent lift of the best-of-N selected perturbation decays quickly with selection-set size m, strongly suggesting overfitting to t
post-training - arxiv:2611.01256 · cs.LGLoRAForge: Co-Designing Post-Hoc Compression and Execution for Efficient Multi-LoRA ServingShenghui Li, Thiemo Voigt
Multi-LoRA serving allows many task-specific adapters to share a common base model, but scaling to large adapter fleets introduces substantial storage, loading, and execution overheads. Post-hoc compression can reduce the memory footprint of trained LoRA adapters, yet smaller ranks do not automatically translate into faster serving. For instance, padding, factor access, and generic kernels can obscure the retained low-rank structure, while independently recompressing thousands of modules also incurs significant preparation overhead. We present LoRAForge, a system connecting batched factor cons
memory - arxiv:2611.01250 · cs.ROShades of Success: Evaluating Manipulation Quality from Object-Centric ConstraintsLorena Maria Genua, Sarvesh Prajapati, Ananya Trivedi, Taskin Padir
As learned manipulation policies become increasingly capable of solving practical manipulation tasks, assessing how well a task is performed becomes a fundamental challenge. While binary labels confirm task success, they cannot distinguish successful executions with different levels of geometric precision. To address this, we introduce SPECTRA, an object-centric framework for quantifying execution quality in manipulation of rigid objects. From a small set of oracle demonstrations generated in simulation, SPECTRA represents object interactions through progress-aligned relative poses. SPECTRA id
manipulationdiffusion policy - arxiv:2611.01245 · cs.LGLearning What to Ask and What to Believe: Distilling a Variational Bayesian Agent into Language ModelsBaichuan Li, Yanling Chang, Alfredo Garcia
Interactive tasks such as eliciting a user's preferences or diagnosing a hidden mechanism require an agent to track its uncertainty over a latent state, ask informative questions and act on the answers. Language models have no explicit mechanism for this, whereas Bayesian information seeking provides one through posterior inference and expected information gain (EIG), and we ask whether it can be distilled into an 8B language model. Our teacher, a lightweight Bayesian agent with a sequential variational belief and EIG query selection, matches or approaches exact and particle references on thre
agentbenchmark - arxiv:2611.01230 · cs.LGReproducing INP-Former: Reconstructing Intrinsic Normal Prototypes for Industrial Anomaly DetectionYunfei Liang, Juan Pablo Osorio Rivera
We reproduce the central claims of INP-Former, a reconstruction-based method for multi-class industrial anomaly detection. INP-Former extracts Intrinsic Normal Prototypes (INPs) from each test image and uses them to guide a decoder that reconstructs normal features, without a stored memory bank. Using the authors' released code, we reproduce the reported performance on MVTec AD and VisA and extend the evaluation to MVTec AD 2, where image AUROC with the default settings is 63.4, compared with 99.7 on MVTec AD. On MVTec AD, a five-seed ablation shows that most of the performance comes from the
memorybenchmark - arxiv:2611.01220 · cs.RODifferentiable Iterative Learning Control for Black-Box Position-Commanded ManipulatorsMohammed Al-Ghaili, Yakoop Qasim, Hassan Almubarak
High-accuracy repetitive tracking is challenging for position-commanded manipulators whose low-level controllers are inaccessible, while constrained model-based iterative learning control (ILC) can require substantial manual parameter tuning. We present Differentiable Iterative Learning Control (Diff-ILC), an online bilevel framework that adapts the parameters of a constrained ILC update between physical robot executions. A closed-loop differential model is identified once from paired smooth perturbations and subsequently kept fixed. At each trial, a constrained quadratic program computes a fe
manipulator - arxiv:2611.01215 · cs.ROVision-Language Models Are Joint Denoisers in Robotic Action GenerationXicheng Gong, Wenxuan Song, Chengze Liu, Qiwei Li +4
Diffusion-based vision-language-action (VLA) models cascade pretrained vision-language models (VLMs) with diffusion action experts, but typically restrict the VLM to providing static conditioning. This separation limits how much pretrained perception and reasoning can guide action generation, especially in out-of-distribution settings. We introduce JADE, a Joint Action DEnoiser that enables the VLM and action expert to jointly refine actions. Our key insight is that structured action latents allow the VLM to read the current noisy action together with visual observations and language instructi
vision-language-actionvlarobotwin - arxiv:2611.01200 · cs.LGGraCo: A Unified Toolkit of Graph Coarsening Algorithms and MetricsAntonin Joly
Graph coarsening aims to reduce a graph to a smaller one that preserves chosen properties of the original. It is increasingly used to scale graph learning to large graphs, as a non-differentiable step applied once before training. Many coarsening methods have been proposed in recent years, each with its own implementation, notation and evaluation. In this work, we introduce GraCo, a unified toolkit of graph coarsening methods. It relies on a single definition: a coarsening is fixed by a partition of the nodes and by the Laplacian the coarsened graph is consistent with. The coarsened graph, and
memory - arxiv:2611.01197 · cs.ROLearning Agile Humanoid Dodgeball with Onboard Monocular VisionFeng Tian, Yulin An, Chenhao Li, Matthias Heyrman +1
Humanoid robots operating in shared, dynamic environments must respond to rapidly approaching objects while maintaining balance. We study dodgeball as a controlled test of this capability, coupling visual threat assessment with whole-body evasion, recovery, and target reacquisition. We develop a system that integrates onboard monocular perception, temporal ball-state and threat-flag estimation, and learned whole-body control. A ball detector supplies bounding-box histories, and a temporal estimator predicts relative ball position and a binary threat flag. Separate dodge/recovery and search/tra
humanoidwhole-body control - arxiv:2611.01193 · cs.LGOnline Autotuning of Algebraic Multigrid via Contextual Bandits and Reinforcement LearningXudong He, Ruiyu Wang, Yilang Ding, Rui Peng Li +1
Algebraic multigrid (AMG) methods are widely used for solving large sparse linear systems, but their efficiency depends strongly on interacting setup and solve parameters. Effective parameter choices typically depend on the problem. We develop a phase-aware online autotuning method for hypre's BoomerAMG to reduce cumulative time-to-solution over sequences of related linear systems. A LinUCB-based contextual bandit selects mixed-type parameters for AMG hierarchy construction using a shared cost model over PDE descriptors and configuration encodings. A recursive LSTDQ controller uses residual an
benchmark - arxiv:2611.01191 · cs.LGArSciKids: Arabic Curriculum-Aligned Datasets for Scientific Experiment and Multiple-Choice Question Generation in Primary EducationTaghrid Housari, Massa Rankoussi, Jafar Abdul Majeed, Wael Tarabishi +2
Despite growing interest in Arabic natural language processing, curriculum-aligned educational resources remain scarce compared with English, and few public datasets for Arabic curriculum content with structured, generation-ready annotations are available, particularly for primary-school science. We introduce ArSciKids, a collection of two Arabic datasets designed for curriculum-aligned educational content generation: (1) a scientific-experiments dataset containing 999 manually curated records from official primary-school science textbooks in Saudi Arabia, Syria, and Jordan, and (2) an assessm
evaluation protocol - arxiv:2611.01190 · cs.ROClothDojo: Learned Data Generation and Benchmarking for Bimanual Garment Flattening and FoldingZilong Huang, Zipeng Ye, Caicheng Wang, Yuchen Xie +1
Garment flattening and folding require robots to identify garment structure during manipulation and adapt their actions as the cloth deforms. We present ClothDojo, a simulation platform that generates adaptive manipulation trajectories at scale across diverse garments and initial states and evaluates policies in closed loop. ClothDojo encodes each garment's canonical mesh coordinates as an NOCS texture, providing canonical surface-position cues through the visual interface of a pretrained vision-language-action model. We fine-tune the model on human demonstrations and corrections to obtain a d
vision-language-actionmanipulationbenchmark - arxiv:2611.01183 · cs.LGSatisfied, Not Solving: Wireheading of Bounded Homeostatic Rewards in Reinforcement LearningChase J. Metoyer
Bounded intrinsic rewards are often assumed to be safe from exploitation: if a signal can yield at most a fixed amount, gaming it should not pay. We test this for a homeostatic satiation reward, in which needs build up and relieving them is rewarding, and give the agent a cheap action that resets its needs. In MiniGrid DoorKey-8x8, PPO agents with this action solved 2 of 80 seeds, against 71 of 80 on the same seeds without it; in KeyCorridorS3R1, 0 of 20 against 20 of 20. Their policies stayed near uniform. The reward does not ask for this: a telescoping argument caps any policy's satiation re
agent - arxiv:2611.01182 · cs.LGFast and Exact Relative-Position Attention via Structure-Aware KernelsDelyan Boychev, Ivo Petrov, Dimitar I. Dimitrov
Relative-position attention mechanisms offer compelling alternatives to RoPE across various transformer architectures, delivering superior performance in domains such as audio. These mechanisms typically modify attention scores by adding content-position or bias-position terms. Unlike RoPE, which encodes position via per-token rotations directly on the queries and keys, relative-position terms generally require pairwise offset-dependent computation or lookups, complicating their integration into fused attention kernels and introducing severe memory, bandwidth, and compute overheads as sequence
memorylong-context - arxiv:2611.01178 · eess.SYRanking the Parameters Shaping Grid DynamicsAhmad F. Taha
Generator and transmission line parameters do not shape the time-domain behavior of a power system equally; a parameter that dominates one dynamic property can be almost invisible to another. This paper studies how the physical parameters of a detailed and general nonlinear differential algebraic equation (DAE) grid model shape the properties that decide whether system states such as frequencies, voltages, and angles can be inferred, observed, controlled, and kept stable. The rankings are conditional as they depend on where grid sensors and actuators sit, on the operating region and the distur
benchmark - arxiv:2611.01168 · cs.LGLightweight Representation Learning for Low Tucker Rank TensorsHiroki Hasegawa, Yukihiko Okada
Low Tucker rank tensors provide compact representations of multidimensional dependencies, but standard learning pipelines can incur substantial computational and memory overhead from redundant intermediate quantities. We reduce this overhead without modifying the underlying representation objective or parameterization. Our approach combines three complementary components: (i) sequential extraction of only task-required spectral components, avoiding unnecessary full decompositions; (ii) factor-wise multilinear contractions without materializing large Kronecker operators; and (iii) compact laten
memorybenchmark - arxiv:2611.01152 · cs.AICapacity Planning for Resource-Constrained Clinical LLM ServingYi Tang, Geoffrey Martin, Yiye Zhang, Yifan Peng
Deploying large language models (LLMs) under a fixed accelerator budget requires jointly deciding how computational capacity is allocated across model sizes and how incoming requests are routed across that capacity. Although prior work has extensively studied LLM routing and serving efficiency, the relative importance of fleet composition and routing under resource constraints remains less understood. We investigate this question using a controlled four-GPU serving environment with heterogeneous clinical NLP workloads spanning short factual verification, biomedical semantic matching, and long-
long-context - arxiv:2611.01145 · cs.AICivicBench: A Benchmark and Memory Framework for Web-Based Civic Data AnalyticsYangyu Wu, Siru Zhong, Haoyuan Liang, Yuxuan Liang
Municipalities publish service requests, geographic boundaries, weather observations, demographic estimates, and operational guidance through independently maintained Web portals. Civic analysis must combine these sources while preserving their schemas, source links, spatial and temporal scopes, and interpretive limits. Tool-using language agents can perform this integration, but existing task benchmarks rarely test whether a computed claim remains supported by the records used. We present CivicBench, a 400-task benchmark over four cities and two years that materializes civic Web data into fro
memorybenchmark - arxiv:2611.01141 · cs.CVGESTALT: Fast, Generalizable Histopathology Image Synthesis from Gene Expression with Pretrained Pathology Foundation ModelsFelix Ho, Alexandros Graikos, Dimitris Samaras
Spatial transcriptomics measures gene expression while preserving where each cell sits in the tissue. The inverse, what a given transcriptomic state looks like, is less explored but equally useful: it lets researchers visualize the morphological consequence of a molecular state and interpret a spatial assay's molecular calls against real tissue structure. GeneFlow, the current state of the art on the Xenium gene expression-to-image benchmark, trains a model from scratch, incurring a high cost on every new target dataset it is applied to. We instead propose GESTALT, which reformulates gene expr
benchmark - arxiv:2611.01140 · cs.AIEMG-Hybrid: A Dual-Engine Elite Memory--Guided Adaptive Optimization Framework for High-Dimensional Feature SelectionSaba Zare, Mohammad Taheri
Feature selection in ultra-high-dimensional, noisy datasets remains a critical challenge due to exponential search space growth, where conventional binary metaheuristics often suffer from premature convergence. To address this, EMG-Hybrid is introduced as a novel three-level hierarchical optimization architecture designed around decoupled operational planes, feedback-driven scheduling, and isolated state refinement. At Level 1, a Dual-Engine Controller monitors population Shannon entropy and empirical fitness velocity, functioning as an architectural master-switch that dynamically routes execu
memory - arxiv:2611.01139 · cs.LGWorld Models Need Not Be Right to Rank Policies Right: Relative Bounds and Audit LimitsAhanaf Hasan Ariq, Hasaan Ahmad
Selecting a reinforcement-learning (RL) policy usually requires additional environment interaction, while learned world models offer cheap but potentially biased evaluations. This raises two distinct questions: can shared model errors preserve pairwise rankings despite poor absolute values, and do they make a small number of real audits more informative? We derive an exact pairwise simulation identity and a relative bound separating common-mode Bellman residuals from policy-specific error. The analysis motivates RAVEL (Relative Auditing of Value Errors in Learned models), a selective policy to
world model - arxiv:2611.01136 · cs.AISIDRec: End-to-End Semantic ID Representation and Modeling for RecommendationZhuohang Jiang, Yongsen Pan, Pangjing Wu, Zihuai Zhao +2
Semantic IDs (SIDs) have emerged as a promising approach for discriminative recommendation by encoding continuous item rep- resentations into compact discrete identifiers. However, existing SID-based methods face two fundamental limitations: semantic mis- match between content semantics and collaborative preferences, and objective mismatch between SID construction and downstream recommendation optimization. To address these challenges, we propose SIDRec, an end-to-end recommendation-aware SID learn- ing framework that jointly models content and collaborative in- formation while aligning SID co
benchmark - arxiv:2611.01135 · cs.AICompile as Hypothesis: A Headroom Typology for Prompt Optimization on Real Reasoning ModelsZhengshuai Yang, Changyun Li
Automated prompt compilers search for or rewrite prompts under the implicit assumption that optimization consistently improves downstream task performance. We evaluate this assumption on frontier reasoning models through rigorous paired, same-day measurements. Across four optimizer families (genome search, profile-guided mutation, GEPA, and ESPO) and eight tasks, structured search yields no statistically significant accuracy gains on strong reasoners. However, a discrete structural genome provides essential architectural stability, preventing catastrophic bimodal seed-collapse (where unconstra
benchmark - arxiv:2611.01133 · cs.CLRethinking LLM Pruning via Test-Time ScalingMingyang Song, Mao Zheng
Structured pruning reduces the size of large language models, often at the cost of lower accuracy. Whether this decline reflects diminished reasoning capacity or greater difficulty in sampling correct solutions is less clear. We examine this question across expert, FFN and layer pruning on four mathematical and coding benchmarks, comparing original and pruned models through pass@k at matched sampling budgets. Even after substantial pruning, some models match or exceed the original model's reasoning coverage despite lower pass@1, and solve problems left unsolved in the original samples. Retaini
benchmark - arxiv:2611.01129 · cs.AIHiGenTrans: Hierarchical LLM Knowledge Transfer for Generative RecommendationZhuohang Jiang, Yuxin Chen, Haohao Qu, ZeFei luo +12
Generative recommendation promises to simplify conventional multi-stage recommender systems by generating item Semantic IDs (SIDs) autoregressively. However, two obstacles limit its effective- ness in industrial settings: interaction-only supervision underuses the semantic knowledge encoded in pretrained language models, while beam-search likelihood ranks token sequences rather than directly estimating item relevance. To address these limitations, we propose HiGenTrans, a three-level framework that introduces LLM knowledge at the input, representation, and decision levels. First, Global Semant
benchmark - arxiv:2611.01125 · cs.LGWhat Does a Safety Label Measure? Auditing Situational-Safety Benchmarks for Multimodal LLMsWayner Barrios, Kenneth Barrios
Situational-safety benchmarks label each scene SAFE or UNSAFE and score a multimodal large language model (MLLM) against that label, usually by judging its free-form answer, and these scores guide deployment. We ask what such a label measures. A label score can come from the scene, the request or the benchmark's own files. We propose the Benchmark Card, seven measured fields that tell them apart, and fill it for six benchmarks. Only MSSBench chat leaks the label through its image files: image size alone reaches 82.3% accuracy, matching or beating six of seven commercial models. Elsewhere the s
benchmark - arxiv:2611.01120 · cs.CVStreamDuo: A Live Commentator and a Research Agent for Unseen GamesZixian Gao, Penghao Luo, Xiangcheng Cao, Lixin Xiu +4
Streaming video LLMs can now commentate live gameplay, but they are trained and tested on the same games. We study cross-game streaming commentary, in which a model trained on one game commentates games it has never seen, and find that current commentators transfer their timing but not their knowledge of the game world. We propose StreamDuo, which pairs a trainable commentator with a frozen research agent that reads the wiki of the game being played. When the commentator sees something it cannot name, it sends the current frames, and the research agent returns a few wiki entries two seconds la
agent - arxiv:2611.01115 · cs.CLCompSkill: Compiling Execution Experience for Skill RetrievalQi Shen, Ziyue Zhu, Yongqing Jiang, Zhen Xia +6
LLM agents increasingly rely on skill documents, yet most retrievers rank them by description alone and ignore what earlier executions revealed, even though a relevant skill can hurt execution. We present CompSkill, which indexes skills with learned 512-d keys and compiles matched same-task comparisons into skill offsets and capped key residuals (KeyOpt), so experience changes which document the executor receives, never its text. Under a logistic outcome model with unknown task difficulty, matched comparisons cancel that difficulty and only discordant ones carry information; on support executi
llm agent - arxiv:2611.01113 · cs.CLCogRSI: Evolving and Auditing Cognitive Reasoning StrategiesYandan zheng, Haoran Luo, Xinliang Zhou, Anhtuan Luu
Can a language model learn better ways to reason about other people's minds by improving its own strategies, and are the strategies it learns actually right? We study this with cogrsi. A grounded reader marks who saw each event in a story and uses a short written principle to settle disagreements between its readings. With a principle written by experts, it beats self-consistency at the same token cost. A self-improvement loop then writes the principle itself. It keeps an update only if accuracy rises by a calibrated margin, and in one variant only if it also passes counterfactual test pairs.
self-improvement - arxiv:2611.01112 · cs.ROAuriga Technical Report: A Task-Agnostic Agent Harness for Zero-Shot Embodied ManipulationChengcheng Wang, Yuheng Shi, Siyu Xu, Yutong Wang +17
Vision-language models can interpret instructions and plan manipulation tasks, yet struggle when acting directly as robot policies. On RoboDojo, a large planning-execution gap separates high-level task decisions from physical execution: removing arm execution raises GPT-6 Astra's success rate from 22.5% under the reference control interface to 76.2%. We present Auriga, a task-agnostic agent harness that closes 68.3% of this gap in zero-shot manipulation without target-task trials or resets. In Auriga, the model makes all task decisions over an explicit task belief and acts through either direc
embodiedmanipulationgraspagentleaderboard - arxiv:2611.01111 · physics.opticsOptimal Fan-Out and Break-Even Routing Loss in Waveguide-Integrated Ge-on-Si SPAD ArraysThammarat Yawisit, Pittaya Pannil
Splitting a waveguide mode over several Ge-on-Si single-photon avalanche diodes (SPADs) relieves dead-time saturation but multiplies dark counts. We quantify this trade-off with the asymptotic count-based information rate, which prices dark counts and dead time in one unit and is attained by an optimally weighted moment estimator. For a non-paralyzable pixel, the information per pre-dead-time signal event peaks at a signal rate $y^*=\sqrt{D(D+1/τ)}$ with exact efficiency $(\sqrt{1+Dτ}-\sqrt{Dτ})^2$, which sets the optimal fan-out and a critical flux above which equal splitting is optimal. We t
benchmarkmach-zehnder - arxiv:2611.01098 · cs.LGMitigating Catastrophic Forgetting in Task-Free Federated Continual Forecasting via Two-Tier Replay under Concept DriftKhaled Hallak, Oudom Kem
Federated time-series forecasting is increasingly important in applications where temporal data are distributed across privacy-sensitive clients, such as environmental monitoring, smart cities, and industrial sensing. A central limitation of current continual-learning protocols is that they usually assume predefined task boundaries, whereas deployed forecasting systems receive an uninterrupted stream and must decide autonomously when adaptation is necessary. This paper addresses catastrophic forgetting in task-free federated forecasting under concept drift. We introduce a drift-aware continual
memory - arxiv:2611.01089 · cs.LGScute: Exact Batched Serving for PEFT Adapters Beyond LoRAThomas Walker, Richard Baraniuk
Existing serving systems for large language models batch requests for many different LoRA adapters on one GPU. However, these systems reject adapters from other parameter-efficient fine-tuning (PEFT) methods such as DoRA, (IA), VeRA and OFT. One solution to this is to convert these adapters to LoRA, but this induces a significant memory and accuracy cost. Instead, we show that, under some modifications, a batched LoRA kernel can serve all of these methods exactly and efficiently. We achieve this by writing each method in one normal form, a LoRA computation with six slots for an adapter's own v
memory - arxiv:2611.01074 · cs.CVGrounding Across Views: Cross-View Referring Reasoning for Visual GroundingHao Gan, Liuqian Wang, Shuang Li, Yufei Gao +4
Cross-view object geo-localization (CVOGL) aims to localize a target in a satellite image from an observation captured under another viewpoint. Existing methods typically specify the query with a click point or bounding box, marking the target region but underusing the semantic attributes and relations conveyed by natural language. Moreover, referring expressions are often viewpoint-dependent, making query-view descriptions difficult to ground directly in the satellite image. In this paper, we formulate cross-view visual grounding (CVVG), extending conventional point- or box-prompted CVOGL to
benchmark - arxiv:2611.01067 · cs.AIRipple: Deadline-Aware CPU-GPU Sharing for Embodied InferenceJiadong Zhang, Xiaosong Ma
Unified-memory edge devices increasingly run LLM/VLM inference alongside time-critical CPU tasks for perception and control. GPU memory traffic can overwhelm these tasks, yet platforms such as DGX Spark expose neither bandwidth counters nor direct bandwidth regulation. We present Ripple, a CPU-GPU co-scheduler that combines model-derived GPU memory volumes with CPU completion feedback to regulate interference. Ripple divides inference into pausable, layer-based units and translates each CPU job's deadline slack into a GPU-byte budget, admitting work within the budgets of all overlapping jobs.
embodiedmemory - arxiv:2611.01043 · cs.AIDependable Agents Need Hard and Fuzzy RulesFrederick Roth
Language models let software read requests in ordinary language, but they keep rules as people do: usually, and less so under pressure. Conventional code keeps rules exactly, but it cannot read a request. We argue that dependable agents need both, because the two kinds of rule cover overlapping and unequal regions of what an agent might do. Code holds the limits that can be stated exactly. Learned fuzzy rules cover the judgment that cannot. We test the hard layer in three domains, purchasing, delegation of authority and data access, on 1,340 episodes written by authors outside our team. Models
agent - arxiv:2611.01042 · cs.AIConstraint Dependency Graph Retrieval for Compliance Detection on Long SKILL DocumentsShuaitao Zhao, Feng Ni, Lichao Ma, Yuchen Wei +5
Large Language Model (LLM) agents carry out business operations under enterprise documents we refer to as SKILL documents, which take the form of standard operating procedures (SOPs) that state the conditions to verify and the actions to take before an operation proceeds. Long, densely cross-referenced SKILL documents exceed the context budget of small deployed auditors. Similarity-based retrieval then brings in irrelevant rules while omitting mandatory prerequisites, and the two-stage audit policy provides no trainable signal for the retrieval stage. We present SkillCDG, a compliance detectio
benchmark - arxiv:2611.01039 · cs.ROpico-WAM: Evaluating Counterfactual Prediction and Planning Error in a Compact World Action ModelKeivalya Pandya
When a World Action Model fails at control, is the bottleneck proposal generation or trajectory scoring? We introduce pico-WAM, a compact model with supplied physical state for tracing policy-dynamics interactions. Using 206 Push-T demonstrations and 15 training seeds, we compare data regimes and search configurations. Counterfactual branch training improves short-horizon cost ranking and enlarges the coverage advantage of policy-restarted over warm-started search by 4.8 percentage points; restart wins under both data regimes. Replacing learned trajectory scores with simulator costs improves c
world model - arxiv:2611.01038 · cs.ROHALO-LP: A Homogeneous All-Shifted Non-Interior-Point Method for Linear ProgrammingJose Rojas, Ana Aguilar, Eusebio Bugarin
This letter introduces HALO-LP, a novel homogeneous all-shifted non-interior-point method for linear programming (LP). The proposed method combines a homogeneous self-dual (HSD) embedding with all-shifted nonlinear complementarity (NCP) functions and proximal primal-dual regularization, enabling arbitrary primal-dual warm starts without requiring strictly interior iterates. The resulting Newton equations admit a condensation based on a positive-definite matrix that can be efficiently factorized using Cholesky decomposition, while the HSD embedding enables primal and dual infeasibility detectio
quadruped - arxiv:2611.01035 · cs.LGSelf-Normalized Inference from Riemannian Stochastic Gradient PathsXiaotian Chang, Qirui Hu, Yi Liu
Riemannian stochastic gradient paths estimate constrained parameters, but confidence regions usually require estimating the Hessian and gradient-noise covariance. We develop self-normalized inference from one path by expressing retained iterates in a common random tangent space. Their average and centered partial sums share an unknown covariance factor that cancels in a Brownian bridge pivot. Under suitable regularity conditions, a joint expansion of the averaged estimator and integrated path controls nonlinear coordinate errors using conditional second moments. The procedure uses fixed-dimens
memory - arxiv:2611.01031 · cs.LGTrouble With Time: Compile Timestamps and First-Seen Date Regression in EMBER2024Edward Raff, Aditya Kapoor
Reliable first-seen dates are important for constructing temporal malware evaluation splits, but this information often comes from external services that are not equally accessible to all researchers. We investigate whether machine learning can instead estimate an executable's first-seen date or determine when its PE/COFF compilation timestamp is unreliable. Using EMBER2024, we evaluate feature-based and raw-byte models across classification, regression, ranking, and latent-reliability formulations. Direct first-seen regression produces errors comparable to a substantial portion of the dataset
manipulation - arxiv:2611.01025 · cs.LGHierarchical Asynchronous Federated Learning over Space-Air-Ground Integrated NetworksKai Wang, Chee Wei Tan
With the emerging 6G paradigm where AI services operate over decentralized and resource-constrained edge infrastructure, performance of distributed learning algorithms remains a significant challenge. This paper proposes HAFL, a Low Earth Orbit (LEO) satellite-assisted hierarchical asynchronous Federated Learning (FL) framework, enabling resource-limited terrestrial clients to collaborate via Ground-Satellite Links (GSLs). Within this framework, LEO satellites serve multiple roles: as edge servers aggregating and broadcasting models, as clients accelerating training, and as high-throughput rel
benchmark - arxiv:2611.01022 · cs.LGJet-QSparse: Accelerating Agentic LLM Inference with 4-Bit Sparse AttentionHaozhan Tang, Hao Kang, Meng Lu, Yecheng Wu +6
Long-horizon reasoning and agentic workloads strain KV-cache capacity and decode attention throughput. We present Jet-QSparse, a training-free method that combines compact NVFP4 cache storage with native FP4 query-key computation on Blackwell tensor cores. A dense scan scores the full key history and selects 128-token blocks for sparse FP16 value aggregation with FP32 accumulation, preserving higher precision in the output-forming step while enabling contiguous tile loads. Calibrated key centering and at most two BF16 pages per sequence and layer address quantization error; FP8 prefill uses a
agenticbenchmarkarena - arxiv:2611.01015 · cs.AIIJev: Order-Invariant Decisions via Iterated Cyclic ReorderingYouzhi Zhang
System One decision models answer a choice question in one forward pass, but their answer depends on the order in which options are listed, and their probabilities carry a prior over answer labels. An invariant ordering operator is any fixed function that returns the same ordered option list for every input order of a fixed option set. Starting from any such order and averaging a listwise readout over its K cyclic shifts removes the input order, but the decision still depends on the chosen cyclic class: those $K$ shifts are only one of the (K-1)! classes, and different invariant operators can
benchmark - arxiv:2611.01011 · cs.AILeap-VLA: High-Rate Robot Control from a Low-Rate VLM BackboneLin Yang, Yunjie Shen, Zheyu Li, Zhenjie Tian +7
Vision-language-action (VLA) models re-encode the full observation through a multi-billion-parameter vision-language model (VLM) at every control step, even though scene semantics evolve at roughly 1 Hz while reactive control requires 20-50 Hz. We present a VLA that decouples the two: a VLM runs as a low-frequency semantic anchor whose cached context is reused across control steps, while a flow-matching action expert operates at high frequency, consuming the cached semantics with a fresh, low-latency visual stream from a compact encoder pretrained on robot video with masked reconstruction, opt
vision-language-actionvlalibero - arxiv:2611.01001 · cs.AIGURU: A Hierarchical Neuro-Symbolic Multi-Agent System for Verification-Aware Automation of Mission-Critical Digital Engineering and OperationsAllan Grosvenor, Anton Zemlyansky, Abdul Wahab, Kyrylo Bohachov +3
An AI can find a safe action faster than any human, but it must never decide what counts as safe. We present GURU, a hierarchical neuro-symbolic multi-agent system for mission-critical digital engineering and operations, built on one partition of authority: learned models propose, rank, and accelerate; a symbolic rules engine checks every consequential action against hard limits before it executes; and validated physics confirms and decides. GURU runs a workflow as a closed control loop, with Navigation, Action, and Planning agents sharing one Common Internal Language and one training factory.
world modelmulti-agentagent system - arxiv:2611.00988 · cs.AIPulse: Long-Horizon Tasks Need Streaming PersonalizationQijia Zhuang, Jialiang Liang, Yuxiang Ren
Large language model agents increasingly undertake complex, long-horizon tasks that benefit from sustained personalized support through long-term memory. Existing memory-augmented methods retrieve and use memory primarily from the initial request, with limited awareness of ongoing execution and emerging personalization needs. Agents may therefore overlook user preferences and project constraints during multistep tasks. We propose Pulse, a streaming personalization method for long-horizon tasks. An independent, lightweight memory agent continuously monitors execution, identifies emerging person
memoryagent - arxiv:2611.00978 · cs.ROGeneralizable robotic manipulation through compositional skill reuseHao Ren, Shuohan Chen, Senyuan Mu, Ruihan Huang +2
Hao Ren, Shuohan Chen, Senyuan Mu, Ruihan Huang, Wanfeng Shang, Yajing Shen
manipulation - arxiv:2611.00957 · cs.ROEvaluating Tactile Foundation Models for End-to-End Imitation Learning in Contact-Rich ManipulationMasaki Murooka, Ryoichi Nakajo, Tomohiro Motoda, Koshi Makihara +3
Tactile sensing plays an important role in imitation learning for contact-rich manipulation by providing direct observations of physical contact with objects. Recently, several tactile foundation models (TFMs) have been proposed to learn general-purpose tactile representations through pretraining on large-scale tactile datasets. However, how these pretrained tactile representations should be leveraged for end-to-end imitation learning remains insufficiently explored. In this work, we integrate multiple TFMs into Diffusion Policy and systematically evaluate their effectiveness for contact-rich
manipulationtactilediffusion policy - arxiv:2611.00955 · cs.ROTask-Specialized Language-Conditioned Models for Fine-Grained Robotic ManipulationChanho Kim, Soyun Choi
Fine-grained robotic manipulation requires instruction grounding, robust perception, and precise contact control. We describe our submission to the MetaFine Open Competition at CoRL 2026. Five independently trained models use a common architecture with frozen DINOv2 and CLIP encoders, trainable language-conditioned feature modulation, and task-specific action heads. We improve demonstrations through contact-aware grasp filtering, instruction-matched data, and expert correction of insertion failures. At inference, task-dependent strategies include photometric test-time augmentation, stochastic
manipulationaction headgrasp - arxiv:2611.00951 · cs.ROWhy Koopman Structure Does Not Help Differentiable World Models for Humanoid Whole-Body TrackingHaizhou Ge, Mengshuo Wang, Yufei Jia, Lei Han +3
We test whether Koopman structure, linear dynamics in a lifted state space, helps when a humanoid whole-body tracking policy is trained by backpropagating through a differentiable world model, replacing the MLP world model of an otherwise unchanged pipeline for simulated Unitree G1 tracking. It does not: return is at best equal (0.823 versus 0.830), one-step error on identical data is 1.3 to 1.6 times the MLP's, and training is about 15% slower. Lifted rollouts diverge or fail to train a policy; of the schemes tested, re-lifting at every imagined step trains best. Re-lifting never uses the lif
humanoidworld model - arxiv:2611.00945 · cs.LGHieraNTP: Structured Risk Representations for Time-Series Anomaly Detection via Hierarchical Next-Token PredictionYujie Li, Zezhi Shao, Boyu Diao, Zhulin An +3
Time-series anomaly detection commonly converts deviations from learned temporal regularities into anomaly scores. This practice can conflate violations of temporal regularity with task-specific anomaly judgments: the same deviation may be anomalous under one task definition and benign under another. Aggregating heterogeneous deviations into a single score before task-specific adaptation can obscure the evidence needed to distinguish these cases. We introduce HieraNTP, a Hierarchical Next-Token Prediction framework that separates self-supervised predictive modeling from task-specific risk inte
benchmark - arxiv:2611.00944 · cs.AISNAPPerf: Software-Aware Tabular Foundation Model for Retraining-Free Performance Prediction of Configurable SystemsAmine Lbath, Massih-Reza Amini, Vadim Okun
Modern software systems expose hundreds of configuration options whose combinations can substantially affect execution time, memory and energy usage. Measuring software performance resulting from a configuration set requires running a realistic workload. Since the configuration space grows exponentially, exhaustive benchmarking is infeasible. Prior work therefore trains models on a small set of measured configurations to predict the performance of the rest. Yet these models are specific to a particular software version, hardware platform, and workload, so changes require new measurements and r
memorybenchmark - arxiv:2611.00941 · cs.LGMarkovTS: Pretrained State-Transition Representations for Zero-Shot Time Series GenerationYujie Li, Zezhi Shao, Jingbo Zhou, Yongjun Xu +2
Across heterogeneous real-world domains, time-series generation supports simulation, model development, and data augmentation by generating sequences that preserve target-domain statistics and temporal dependencies when real observations are scarce or costly. However, existing methods predominantly address in-domain generation, limiting their reuse in unseen domains. We study reference-conditioned zero-shot time series generation: target domains are unseen during pretraining, generation uses a single target-domain reference sequence, and no model parameters are updated. MarkovTS characterizes
benchmark - arxiv:2611.00926 · cs.CVUI-UG 2: Advancing Mobile UI Understanding and Generation with Fine-Grained Supervision and Reinforcement LearningHao Yang, Shuoyang Liu, Xiaoyu Lin, Zhou Fang +5
Mobile UI understanding and generation require both fine-grained component perception and reliable structured synthesis. Inspired by the unified direction introduced by UI-UG, we present UI-UG 2, a compact 4B-parameter multimodal large language model (MLLM) that advances mobile UI Understanding and Generation with broader task coverage, a front-end-aligned taxonomy, and online render-aware reinforcement learning. UI-UG 2 constructs a multi-source understanding corpus from RICO, VINS, Android in the Wild, and Android Control, annotated with an 80-category two-level UI taxonomy aligned with fron
benchmark - arxiv:2611.00923 · cs.AIIllusion: Safeguarding Topology Confidentiality in LLM Multi-Agent Systems via Behavior-Preserving ObfuscationYiu-Fai Cheung, Hongzhi Guo, Miao Pan, Irwin King +1
The communication topology of Large Language Model Multi-Agent Systems (LLM-MASs) governs how information propagates among agents and encodes critical domain knowledge and expert workflow design. It is valuable intellectual property and plays a decisive role in system utility and safety, yet remains vulnerable to inference attacks that stealthily reconstruct it from agent outputs. Existing defenses detect or mitigate such attacks but do not proactively conceal the topology itself. We propose \textit{Illusion}, a proactive topology obfuscation defense for LLM-MAS. \textit{Illusion} injects fake
agentmulti-agentagent system - arxiv:2611.00920 · cs.LGWhat matters in classical machine learning for non-intrusive load monitoring? A factorial benchmark on REFIT with validation on unseen housesZablon Ombiro
Published classical machine-learning results for appliance ON/OFF classification in nonintrusive load monitoring (NILM) range from F1-type scores of 0.53 to 0.98, and the sources of this spread have not been separated. This study separates them with a factorial benchmark on REFIT: seven models, two feature sets and four class-imbalance strategies for five appliances (270 fitted configurations), each evaluated by pooled nested cross-validation and on houses unseen during training, with robustness checks. An analysis of variance shows that the appliance explains 38% of the variance in F1, the fe
benchmark - arxiv:2611.00917 · cs.LGBenchmarking Classical Machine Learning Models for Shading-Fault Detection in Rooftop Photovoltaic Systems Using Real Sensor DataZablon Ombiro, Bernard Shibwabo
Most published comparisons of machine-learning classifiers for photovoltaic (PV) fault detection are validated on simulated faults. We compare seven classical and ensemble-tree classifiers on the Réunion dataset, a 4 kW domestic rooftop plant on which shading faults were physically staged and logged. Readings within a staged episode are near-duplicates, so every train/test split is grouped by episode. A 20,800-row subsample that retains all 17 fault episodes was evaluated with three repeats of five-fold group-aware nested cross-validation, with synthetic minority oversampling (SMOTE) applied i
benchmark - arxiv:2611.00911 · cs.AIS1CogBench: What System One Decision Models Can and Cannot Do on Cognitive ExperimentsDengzhe Hou, Kazunori D Yamada
Language models are no longer used only to generate text. Software also asks them for decisions, in routing, guardrails, and automated gates. System One decision models, also called typed decision models, such as Jev and Laya, return one fast decision over the legal options, leaving no text to inspect. However, a fast, well-formed decision does not show that the model used the evidence. Cognitive experiments can test this, because they vary task demands and measure how behavior changes, yet existing cognitive benchmarks evaluate language and multimodal models. We introduce S1CogBench to ask wh
memorybenchmark - arxiv:2611.00908 · cs.CVZOPGA-KD: Zeroth Order PGA Guided Data Free Knowledge DistillationTuhin Subhra Mukherjee, Tiya Rose Pulikunnel, Vaidhav Adit, Devanand Thanissery +1
Knowledge distillation (KD) enables the transfer of learned representations from a large teacher network to a small student network. In foundational KD, the student network uses the original training data of the teacher model to accomplish this task. However, due to privacy concerns, the original training data is often not available. To overcome this reliance on original training data, several data-free knowledge distillation methods have been developed. Some of these methods operate in the white-box setting, and a few methods generate synthetic data by training a generator network. White-box
benchmark - arxiv:2611.00907 · cs.LGBenchmarking Machine-Learning Classifiers for Neoantigen Immunogenicity Prediction and the Limits of SHAP InterpretabilityRamon Gonzalez
Personalized cancer vaccines require prioritizing, among hundreds of candidate mutant peptides, the few that will induce a T-cell response. This study asks what general-purpose machine learning achieves on that task, and whether the features such models identify as important can be trusted. We make three contributions. The first is a benchmark of sixty models: twelve supervised classifier families, each in five configurations, evaluated on the Immunogenic Tumor Specific Neoantigen database (ITSNdb; 199 experimentally validated neoantigens, 129 immunogenic and 70 non-immunogenic). All sixty are
benchmark - arxiv:2611.00899 · cs.CVFilmTogether: Cinematic Video Generation from Low-Cost InputsShiying Zhang, Xiaobin Hu, Qianchi Zhang, Jiaqi Zhao +3
The widespread adoption of mobile devices has democratized video recording, but a stark aesthetic divide persists between casual, low-cost captures and high-cost cinematic productions. Bridging this gap requires upgrading heterogeneous production elements, such as lighting, specialized costumes, and materials, without altering the core layout, identity, and temporal motion of the original footage. However, this task is heavily bottlenecked by two important challenges: 1) Lacking a high-quality paired training data specifically designed to handle complex low-to-high appearance shifts. 2) Ignori
memory - arxiv:2611.00898 · cs.ROTrading Transparency for Stability: Friction-Cone-Constrained Bilateral Teleoperation for Opening Doors of Unsecured FurnitureCheng Zhou, Jing Li
Opening the door of unsecured furniture loads the frictional contacts at its supports; too large a pull slides or tips the object. A friction-cone constraint on the commanded wrench removes this failure mode but costs haptic transparency. We quantify the trade-off: an admissible wrench set built from contact friction cones and support-polygon tipping margins is shrunk for robustness to bounded mass and friction errors, and any command in the shrunken set provably leaves the furniture static (Prop. 1). We call this map the robust friction cone projection (RFCP). RFCP and its haptic guard admit
teleoperationgrasp - arxiv:2611.00896 · cs.ROVerifying Test-State Identity in Paired Evaluation of Vision-Language-Action ModelsRafiqul Islam
Simulation-based VLA evaluation often treats matching task, seed, and episode identifiers as sufficient evidence that two policies faced identical test conditions. Under specific vectorized reset and autoreset conditions, different termination timing can change subsequent initial-state assignment, so the same episode index need not mean the same test state. We present an evaluation-integrity framework for paired comparison - invariants I1-I5, provenance logging, optional scene fingerprints, and an automated checker - and a controlled LeRobot 0.6.1 LIBERO harness study with emulated controllers
vision-language-actionvlaliberoleaderboard - arxiv:2611.00892 · cs.CVLearning to Group and Order: Cross-Instance Self-Supervised RL for Vision-Centric MLLMsSongming Yang, Hong-Tao Yu, Xiu-Shen Wei
Reinforcement learning (RL) post-training has become an important paradigm for improving multimodal large language models (MLLMs), especially with automatically verifiable rewards. Self-supervised visual pretext tasks provide such rewards without human annotations. However, existing methods mainly exploit within-instance structure, offering limited supervision for cross-instance discrimination and fine-grained comparison. We introduce Group-and-Order Self-Supervised Reinforcement Learning (GO-SSL), a cross-instance jigsaw framework for vision-centric MLLM post-training. Given two visual instan
post-trainingbenchmark - arxiv:2611.00884 · cs.CLWhen 200 OK Is Not an Answer: Silent Fetch Failures in Agentic Web Content AcquisitionShuyang Zhang
Web agents and answer engines fetch pages at question time, and liveness checks of cited sources treat an HTTP 200 response as success. We measure how often that assumption fails and what it does to measurements built on fetched pages. We fetched 10,307 URLs in three days with a polite static client and a rendered browser client: pages from 1,507 sites in five strata, random non-existent paths on every site, and the 4,795 URLs cited by expert-verified claims in ExpertQA. We define a taxonomy of what a 200 response contains, label the static 200 responses automatically, and check the labels aga
agentic - arxiv:2611.00883 · cs.CLRandom Is Almost Enough? Stable and Low-Cost Benchmark Compression via BestRandGeorge Ling, Shanshan Zhong, Lijin Yang, Hao Yang +2
Benchmark compression reduces the often prohibitive time and cost of model evaluation by evaluating models on a small benchmark subset that closely matches full-benchmark results. To quantify estimation error, we derive an accurate analytical approximation to the error distribution when estimating full-benchmark scores from a subset, showing that most subsets yield small errors, while a few produce much larger errors, resulting in a long-tailed distribution. This means random sampling alone can achieve surprisingly strong performance with high probability, but remains unstable due to rare larg
vlaagentagent benchmarkbenchmark - arxiv:2611.00871 · cs.AIShiftRAG: Bypassing the Textual Bottleneck via Decoupled Learning and Continuous Soft TokensYoumin Ko, Jihong Jeong, Hyunjoon Kim
Despite advances in retrieval-augmented generation (RAG), suppressing hallucinations during complex reasoning remains a persistent challenge. We identify that existing RAG pipelines suffer from a textual bottleneck: by passing only raw text to the generator, they discard discriminative signals of the retriever. Furthermore, recent attempts to unify retrieval and generation within a single LLM induce severe objective conflict and exacerbate the rank collapse of decoder-only architectures, degrading performance on both tasks. To address these limitations, we introduce ShiftRAG, a decoupled RAG f
retrieval-augmentedragrag pipeline - arxiv:2611.00870 · cs.ROA Reproducible Reconstruction and Evaluation Framework for Photorealistic Lunar Analog Scene Reconstruction Toward Digital TwinsAntar Mazumder, Ryan Hartzell, Luke Boyd, Alexander Robinson +3
In this paper, we present a reconstruction pipeline that empirically quantifies the photorealistic scene reconstruction accuracy from both human-controlled and rover mast camera monocular feeds against physical ground truth as a function of Lunar traversal effects (e.g., motion blur, haze due to lofted regolith, trajectory bias, and sparse views). First, we benchmark six variants of 3D Gaussian Splatting (3DGS) across three Lunar analog scenes on held-out test sets using standard photometric metrics. We then establish how under traversal effects, this test-set-specific Novel View Synthesis (NV
benchmarkevaluation framework - arxiv:2611.00852 · cs.LGAdaptive-Policy Distillation: Knowledge Injection under Information AsymmetryHojun Cho, Kang Eun Jeon, Jaegul Choo
On-policy distillation (OPD), which uses reverse-KL supervision on student rollouts to transfer a context-conditioned model's behavior into student model parameters for closed-book inference, has shown strong results. However, we find that this success does not translate to knowledge injection: across two document-QA benchmarks, OPD injects document knowledge far less efficiently than simple off-policy distillation. We trace this reversal to two failures that are structural, yet sparse across token positions. Reverse KL weights its gradient by the student distribution and therefore barely lift
benchmark - arxiv:2611.00848 · cs.CVToward Counterfactual Learning for Selective Re-Examination in Agentic Visual SearchAo Chen, Shengchao Chen, Yan Luximon
Visual search over high-resolution images requires a vision-language model (VLM) not only to determine where to inspect, but also whether another observation is warranted before committing to an answer. Existing prompting, demonstrations, and outcome-based reinforcement learning can encourage re-examination, yet do not directly supervise whether it is beneficial at a particular search state. We introduce ReSee, a benchmark of re-examination decisions comprising 3,227 visual-search trajectories and 8,827 explicit commit-or-re-examine decisions, with counterfactual adjudication that distinguishe
agenticbenchmark - arxiv:2611.00841 · eess.SYRobust Data-Driven Output Feedback for Systems With Unknown Input DelaysNgoc Tuan Dinh, Ba Huy Nguyen, Alexey Peregudin, Egor Dogadin
In a history-feedback realization, an input delay increases memory without increasing the number of unknown physical coefficients. We use this structure to design output-feedback controllers for discrete-time systems with one input, one output, and known undelayed order, using noisy input-output data. For a known delay, we derive a necessary and sufficient linear matrix inequality for common-quadratic stabilization of all models consistent with the data. Optimizing its equation-error bound gives a robustness margin for the measured record. For an unknown constant delay in a known interval, we
memory - arxiv:2611.00838 · cs.AICan Agentic AI Undermine Its Own Assurance Loop? Closed-Loop Dependency Assurance for Cognitive Network Digital TwinsAhmed Hussain, Konstantinos Kalogiannis, Salahuddin Salahuddin, Hexu Li +1
Agentic Artificial Intelligence (AI) and Network Digital Twins (NDTs) are increasingly combined to form Cognitive Digital Twins (CDTs) that translate operational intents into network changes and support closed-loop management. However, a network change may achieve its intended objective while disrupting the monitoring or recovery mechanisms needed for subsequent decisions. For example, a service may be rerouted while its latency-monitoring traffic continues along the original path, producing measurements that no longer reflect the service's performance. We propose Closed-Loop Dependency Assura
agentic - arxiv:2611.00828 · cs.LGAnswer Correction with Verifiable Rewards: A Gemma-Family Study of Multi-View Driving QAT N Rickesh, Pranav Vinodh, Sowmya Kamath S
Multi-camera driving question answering provides a discrete test of vision-language models, but post-training gains can reflect changed selected answers or improved answer extraction. We investigate this distinction within the Gemma model family using archived, task-linked WaymoQA evaluations. A zero-shot diagnostic on 5,519 test questions spans effective-small, mixture-of-experts, unified, and dense checkpoints, with serving differences disclosed rather than interpreted as architecture effects. The strongest controlled evidence is a matched-harness Gemma-4-12B base/GRPO pair: accuracy rises f
post-training - arxiv:2611.00822 · cs.LGAdKnob: Ad Intensity Control and Labeling for LLM-Native AdvertisingWoo Jae Kim, Seongho Keum, Joonsung Jeon, Suhyeon Ha +2
Large language model (LLM)-native advertising has emerged as a viable monetization channel as LLM service providers seek sustainable revenue streams. However, integrating native ads into LLM responses introduces two largely understudied challenges: (1) controlling ad intensity -- how prominently advertising content appears -- and (2) ad labeling -- identifying which segments of the generated text constitute advertising. We propose AdKnob, a framework jointly addressing both challenges. For ad intensity control, we introduce new ad control tokens without pretrained semantic associations and ali
evaluator - arxiv:2611.00819 · cs.LGODC: Orthogonal Drift Correction for Improved Text-to-Image Semantic Alignment at InferenceKasra Kamali, Basim Azam, Tuan Duong Trinh, Jean Honorio +1
Text-to-image models have achieved remarkable success in generating high-quality images from textual descriptions. However, they often struggle with "semantic drift," where the generated output fails to precisely align with complex or nuanced text prompts. While recent approaches have attempted to address semantic errors regarding attribute binding or object presence, there remains a gap for a more holistic method that addresses these issues by directly refining the text embeddings of the initial user prompt. In this work, we introduce Orthogonal Drift Correction (ODC), an inference-time guida
benchmark - arxiv:2611.00810 · cs.CVUniCaMo: Unified Camera and Motion Control for Video Diffusion ModelsLong Vu, Tan Ngo, Animesh Karnewar, Amir Habibian +4
We present \textbf{UniCaMo}, an unified framework for joint control of camera and object motion in image-and-text-to-video diffusion models. Given a reference image, text prompt, sparse 3D object trajectories, and a 3D camera trajectory, UniCaMo constructs motion-consistent input noise that directly encodes the desired motion. Our 3D-grounded noise construction combines reference-derived noise propagation with a spherical Gaussian noise representation for newly revealed regions, providing a simple and effective motion prior without requiring reconstruction of unseen scene geometry. UniCaMo req
benchmark - arxiv:2611.00806 · cs.LGStreaming Virtual Metrology via Lightweight Adaptation under Temporal Distribution ShiftsDoyun Choi, Cheonwoo Lee, Younghyun Han, Kiwook Han +4
Virtual metrology (VM) aims to estimate costly or delayed metrology variables from heterogeneous sensor signals and in-line measurements, and it acts as a critical component of semiconductor manufacturing. While recent data-driven models have improved VM accuracy, most are trained in batch settings and assume stationary data distributions. In practice, however, production environments evolve continuously due to equipment degradation, process adjustments, and maintenance events, making conventional batch-trained VM models difficult to maintain under temporal distribution shifts. In this work, w
online learning - arxiv:2611.00796 · cs.CLSteer, Don't Tell: Breaking Reasoning Plateaus with Reachability-Aligned Self-DistillationZijian Wu, Lingkai Kong, Lin Yang, Haiteng Zhao +4
Post training for large reasoning models remains insufficient on challenging problems, whose correct trajectories are hard to sample. Though prior works address this difficulty by integrating supervised fine tuning on expert trajectories with reinforcement learning, such external experts are not always available. Existing self improvement methods instead use reference solutions to create a privileged self-teacher. Yet privileged teacher correctness does not imply student reachability. Solution conditioning shifts the teacher distribution toward actions unlikely under the unprivileged student,
self-improvementpost trainingiterative refinement - arxiv:2611.00795 · cs.AIBridgeDock: Taming Generative Co-Folding Prior for Molecular Docking with Diffusion BridgeShikun Feng, Zipeng Yang, Qiuyi Li, Mingze Yin +3
Molecular docking aims to predict the three-dimensional structure of a protein-ligand complex; in the conventional structure-based setting, it is conditioned on an input protein structure and ligand, often starting from an apo receptor in flexible docking or a holo receptor in rigid docking. Recent co-folding models exemplified by AlphaFold 3 achieve strong accuracy in protein-ligand complex prediction, but they typically generate complexes from sequence-derived inputs and therefore do not directly use apo structures as structural priors in docking, leading to a larger conformational search sp
benchmark - arxiv:2611.00791 · cs.CVPhoneValue-Bench: A Multimodal Benchmark and Datasets for Consumer Smartphone Recommerce ValuationZihan Deng, Jun Gao
As consumer smartphone recommerce scales, buyers must quote binding repurchase prices from seller-uploaded multi-view photos and diagnostic screenshots before devices reach the warehouse. Accurate remote valuation requires fusing noisy cosmetic evidence with identity, battery, lock, and channel risk from screenshots and staff forms, since an error in any modality changes both grade and payout. However, existing multimodal resources mainly target general VLM competence, screen reading, or factory defect detection, and largely overlook staff-aligned trade-in fields, dual appearance/staff ground
benchmark - arxiv:2611.00788 · cs.LGTight Stochastic Sequential Covers for General VC Classes with Unknown i.i.d. FeaturesChanglong Wu, Wojciech Szpankowski
We study hybrid online learning with stochastic features and adversarial labels when the feature distribution is unknown. Prior work gave stochastic sequential covers of log-size $O(d\log^2 T)$ for a general VC class and logarithmic bounds only under stronger star-type structure. We show that finite VC dimension alone is enough: every binary class of VC dimension $d$ admits a distribution-blind exact cover with log-size $O(d\log T+\log(1/δ))$, matching the worst-case $Ω(d\log(T/d))$ dependence on the horizon. The key ingredient is a new local cube weight that counts Boolean subcubes incident t
online learning - arxiv:2611.00782 · cs.AIOne Offset, Two Directions: Baseline Adjustment via Subgroup OffsetsJiaying Zhou, Jiachen Liu, Qingchao Chen
Group Relative Policy Optimization (GRPO) estimates advantages against the within-prompt mean reward, a baseline shared across the group. In practice, rollouts within a prompt are not always homogeneous, differing in their solution strategies and reasoning trajectories, along which they may group into subgroups. We further find that these heterogeneous subgroups may exhibit distinct reward distributions. The global baseline, however, collapses them onto a single credit axis, where the credit reflects neither a rollout's standing within its subgroup nor the subgroup's standing relative to the r
benchmark - arxiv:2611.00779 · cs.AIWho Schedules Agent Memory? A Sidecar Approach to Long-Horizon TasksQijia Zhuang, Rui Liu, Yuxiang Ren
As large language model agents increasingly undertake complex, long-horizon tasks, memory has become an important component for supporting sustained decision-making and action. Existing methods primarily focus on memory storage, updating, and retrieval, but often overlook how memory scheduling affects task execution: the main agent must identify memory needs, initiate retrieval, and wait for results while advancing the task, increasing both scheduling burden and execution time. To address this problem, we propose MemSidecar, which decouples memory scheduling from task execution through two ind
memoryagent memoryagent - arxiv:2611.00774 · cs.AIDecision-Centric Research: A Calibrated Decision Model and Benchmark for Scientific AgentsShaoning Zeng, Wen Huang, Xinzhong Zhu, Yi Duan +2
Scientific agents make consequential choices about evidence, next actions and research strategy, but choices embedded in generated text are difficult to evaluate independently. Decision-Centric Research (DCR) is introduced to make individual choices the common unit of supervision, prediction and evaluation through typed predictions over an explicit Research State. DCR defines 24 decision types, ten of which are instantiated in the Research Decision Benchmark (RD-Bench) of 2,000 rule-generated cases. The Decision-Centric Research Model (DCRM) shares a candidate-token readout across types to pro
benchmark - arxiv:2611.00773 · cs.AIDesigning Informative Actions: Action Construction for Long-Horizon Research AgentsAlkut Arxidin, Shaoning Zeng, Jianhang Zhou, Jianping Gou +1
Large language model agents support automated scientific research by connecting literature analysis and planning with experimentation. Evaluating which action an agent selects does not establish which experiment its instruction specifies, because omitted interventions or replication counts leave consequential choices to the executor. We introduce a reading-set framework for evaluating action construction under a common cost rule. Across 30 synthetic states, complete-design requests increase affordable efficiency by 465.98% and 327.91% relative to the action-name baseline in the two model famil
agent - arxiv:2611.00769 · cs.CVDramaVQA: Probing Detailed Understanding of Ultra-Long TV DramasXinyue Huo, Pengfei Chen, Jiannan Ge, Kaiwen Duan +4
Despite recent progress in long-form video understanding, current benchmarks mostly focus on short clips or hour-scale videos. This leaves a major gap in evaluating ultra-long video understanding, where videos often last tens of hours. To bridge this gap, we introduce DramaVQA, a large-scale benchmark built upon long-form TV dramas. Containing approximately 20K question-answering (QA) pairs, the scale of DramaVQA exceeds previous long-video datasets, such as Video-MME and LVBench, by an order of magnitude. Our benchmark has two main features that make it stand out from existing datasets. First
agenticbenchmark - arxiv:2611.00768 · cs.CVDADKC: Distribution-Aware Adaptive Differentiated Knowledge Consolidation for Cloth-Hybrid Lifelong Person Re-identificationYan Zhang, Minghao Long, Meiqi Wan, Yan Yang +3
Lifelong Person Re-identification (LReID) aims to match the same person using sequentially collected data. In practical long-term deployments, however, the model must handle both domain shifts across cameras and scenes, and irregular clothing changes of the same identity. This gives rise to the Cloth-Hybrid Lifelong Person Re-identification (CH-LReID) task, which demands balancing cloth-relevant and cloth-irrelevant knowledge in streaming data. While existing LReID methods have explored replay-based or distillation-based anti-forgetting strategies, they typically fail to sufficiently model sta
benchmark - arxiv:2611.00762 · cs.AIHiPaper: A Human-AI Collaborative System for Academic Literature Triage and Scholarly SensemakingYe Jia, Pangjing Wu, Hayley Y. W. Gu, Zhuohang Jiang +5
Scholarly sensemaking requires researchers to iteratively retrieve, collect, and synthesize literature as their understanding evolves. However, existing scholarly tools provide only partial support for this process. Recent artificial intelligence (AI)-based systems generally rely on opaque submit-and-wait pipelines that it users' ability to inspect, revise, and validate intermediate outputs. In this study, we present HiPaper, a human-AI collaborative literature review system informed by a formative study with academic experts. HiPaper organizes scholarly sensemaking as a three-stage interactiv
human-in-the-loop - arxiv:2611.00758 · cs.LGOld Is Not False: Valid-Time Memory for Interactive Video World ModelsZiming Wang, Chuyue Gong, Hongliang Cao, Jingmin Zhu +3
Interactive video world models draw each frame from a memory bank of earlier frames stamped with their capture times. With frames, weights and noise fixed, reversing the timestamps of two views flips the state that generators of two families draw: they take the newest timestamp as the truth. That reading is right as long as states never return. After an agent undoes an action, the newest view of a region is outdated and the only view of the present is older; memories for changing worlds delete that older view, and memories that keep it hand the generator the past. Temporal databases separate w
world modelmemoryagent - arxiv:2611.00756 · cs.ROSceneBound: How Much Should a Streaming 3D Reconstruction Model Remember?Ziming Wang, Chuyue Gong, Chutian Wang, Jingmin Zhu +3
Streaming 3D reconstruction models keep past video frames in a recurrent state, fast weights from test-time training, or a key-value (KV) cache. Each written frame shifts the coordinate frame of later cameras or occupies GPU memory. Training-free rules decide what to write from a signal of the current frame. How much they write accounts for their mean accuracy: spending the update budget of a gate at uniformly spaced frames, with no signal, matches nine published gates, and the budget bounds the drift of the state wherever the writes fall. We propose SceneBound, a training-free rule that write
memory - arxiv:2611.00755 · cs.CLDuSE-Bench: Benchmarking Joint Reasoning over Evolving Dialogue and DocumentsGuanjie Cheng, Jianfeng Gao, Yufei Tong, Xinkui Zhao +1
Persistent Web information assistants must combine evolving document facts with user requirements established and revised across sessions. Yet systematic evaluation of how changes in both sources affect response requirements and evidence attribution remains limited. To address this gap, we introduce Dual-Source Evolution Benchmark (DuSE-Bench), a benchmark comprising 1,000 tasks across five scenarios, including 700 whose required premises originate in both dialogue and documents. An evolving Support Graph guides task construction and preserves links from updated facts and reused conclusions to
memorybenchmark - arxiv:2611.00752 · cs.AIReading Between the Chapters: A Graph-Free RAG Rebuilt With AI and Human FeedbackDipto Sumit, Syeda Fabiayee Jahan, Farig Sadeque
Many questions about a book can only be answered by connecting passages chapters apart, such as a cough heard three times across a horse's life. We introduce CrossChapterQA, 1,853 such questions over 14 public-domain books. Each question lists the two to six chapters it needs (spanning half the book at the median) and the relation linking them, cites evidence from each, and has two independent reference answers graded blind. We use it to build LadderRAG, which indexes a book as passages, summaries and book-level cards for about $0.41 and answers a question for under a cent. Its first version f
ragrlhfrlaif - arxiv:2611.00749 · cs.LGPeptSol: A Translational Benchmark for Protected-Peptide Solubility in Drug ManufacturingAlexander Arjun Lobo, Marc Saouda, Urvi Awasthi, Leonid Zhukov
Liquid-phase peptide synthesis (LPPS) offers a scalable route to peptide manufacturing, but route development still depends on empirical solvent screens because each protected intermediate must be soluble for reaction and recoverable for purification. The protecting groups, terminal modifications, and noncanonical residues that govern this behaviour are poorly represented in existing peptide property benchmarks, limiting the ability to develop and rigorously evaluate molecular representation models for LPPS-relevant prediction tasks. We introduce PeptSol, an open benchmark of 1388 hand-curated
benchmarkevaluation protocol - arxiv:2611.00746 · cs.AIResMamba: A Dual-Path Convolution-Mamba Model for Music Genre ClassificationYingyu Zhang, Pu Qu
Music genre classification remains a fundamental yet challenging task in music information retrieval, requiring models that capture local time--frequency patterns and long-range temporal dependencies. To address this, we propose ResMamba, a hybrid architecture that augments a ResNet-style four-stage backbone with Mamba blocks. Building upon the selective structured state space sequence model, ResMamba first adapts the selective SSM into a 4D-compatible residual Mamba block that captures global dependencies at linear complexity. We then design the ConvMamba dual-path block, which combines this
benchmark - arxiv:2611.00734 · cs.ROMOST: Memory of Object States and Task Progress for Long-Horizon Mobile ManipulationDiqi He, Jinbang Guo, Yijun Shen, Zhiwei Yu +4
Long-horizon embodied manipulation requires policies to act on information that is no longer visible. Modified objects may leave the field of view, while completed subtasks continue to constrain future actions. Existing approaches extend temporal context through recurrent state updates, test-time training, or storage and retrieval of past observations. As horizons grow, these mechanisms can still leave the policy to infer task state from long, redundant, and viewpoint-dependent visual experience. We argue that long-horizon agents benefit not merely from remembering more, but from remembering t
embodiedmanipulationmemorybenchmark - arxiv:2611.00723 · cs.AIArchitecting Fairness-Aware LLM InteractionsAlessandra Parziale, Valeria Pontillo, Fabio Palomba, Patrizio Pelliccione
Large Language Models (LLMs) are increasingly used as decision-support tools but may generate stereotypical, discriminatory, or unfair responses. As LLMs become embedded in decision-critical software systems, fairness can no longer be treated as a downstream evaluation step, but must become an architectural concern of the interaction itself. Yet current fairness approaches operate as standalone evaluation techniques, lacking a reusable mechanism to verify and mitigate violations at runtime. We propose a fairness-assurance architecture grounded in three reusable design patterns, instantiated as
multi-agentagent framework - arxiv:2611.00722 · cs.CVSAVIER: Source-Anchored Visual Inspection and Evidence-Guided Repair for Vietnamese Cultural Text-to-Image GenerationGia-Huy Thai, Duc-Tien Khong, Trung-Nghia Le
We present SAVIER, a training-free, one-pass framework that repairs the depiction of a named Vietnamese cultural entity in a generated image while preserving the scene details requested by the user. The problem is hard because a plausible scene can miss the cues that identify the entity, and prompt-level refinement never inspects the image it produces. SAVIER starts from a Culture-TRIP expanded prompt; a Curator separates source-linked cultural cues from prompt-preservation requirements, a prompt-blind Observer describes the first image, and a Refiner turns the discrepancies into at most three
agent - arxiv:2611.00718 · cs.ROSTEER: Route-Aware Adaptive Reasoning for Autonomous DrivingYangang Zou, Pei Liu, Nan Song, Bozhou Zhang +5
Vision-language action (VLA) models excel for end-to-end autonomous driving, yet their tendency to over-invoke chain-of-thought (CoT) reasoning mirrors a human cognitive pitfall: overthinking. Just as deliberate reasoning can slow or mislead human judgment in routine tasks, excessive CoT in VLA incurs unnecessary compute overhead and can paradoxically degrade planning performance. Recent research tackles this via adaptive reasoning, which dynamically modulates inference complexity by selectively triggering CoT for complex scenes while responding directly to routine ones. While prior adaptive m
vla - arxiv:2611.00713 · cs.LGAdversarial Robustness Is Not Zero-Day Generalization: A Controlled Evaluation of Flow-Based Intrusion Detectors on Corrected CIC-IDS2017Abdulhamid Zaro
Synthetic variants of known attacks, including feature-space adversarial perturbations, are sometimes used as stand-ins for unseen ("zero-day") attacks when evaluating machine-learning network intrusion detection systems (NIDS). We test whether this proxy holds. On the corrected CIC-IDS2017 release we train Random Forest, XGBoost, a 1D-CNN and an FT-Transformer and compare, at a fixed 1% false-positive rate over five seeds, detection under constrained perturbations of known attacks (gradient-based and black-box attacks on 42 attacker-controllable features) with leave-one-attack-family-out (LOA
benchmark - arxiv:2611.00689 · cs.AIGetting a Grip: Operating Human-Agent Software Teams Through Governed MessagesMichael Granitzer, Kanishka Ghosh Dastidar, Andreas Einwiller, Jelena Mitrović +1
Software is increasingly developed by teams in which several people direct many coding agents from different vendors. Such teams must establish who may request what from whom, which messages start work, which decisions bind, and who caused a change. We propose to govern a human-agent team through its message protocol and present yoars, a relay in which humans and agents are authenticated principals that exchange typed, addressed messages in rooms holding goals, decision rules, routing and an attributable, append-only record. This report gives the architecture, the full controlled study and the
agent - arxiv:2611.00686 · cs.LGGenerative Forecasting from Coarse-Grained Observations: When Does Cascading Help?Alexandra Spitzer, Giovanni Abati, Lorenzo Baldassari, Till Muser +2
We study probabilistic forecasting of dynamical systems such as fluids and waves from coarse observations. Since the unseen fine scales influence the dynamics, the same coarse initial state can evolve into a whole distribution of futures. Generative forecasters may produce visually plausible samples, but it is hard to tell whether they sample the right conditional distribution, since trajectory data contain a single future for each initial state. We address this problem by generating reference data in a hierarchical way: for each coarse initial state, we generate many different fine-scale comp
benchmark - arxiv:2611.00684 · cs.CVDensity-Aware Calibration for Resource-Efficient Medical Vision-Language ModelsTien-Dat Dam, Huu-Thuan Huynh, Minh-Triet Tran, Trung-Nghia Le
Medical vision-language models (VLMs) show promise for clinical image understanding, but their computational and memory requirements hinder deployment on resource-constrained clinical hardware. Post-training quantization (PTQ) offers a practical compression strategy without retraining, yet its effectiveness depends strongly on the calibration data. We propose CoreCalib, a training-free and quantizer-agnostic framework for constructing compact calibration sets that preserve the target medical VQA distribution. CoreCalib stratifies data by clinical question class, identifies representative visua
memorypost-training - arxiv:2611.00681 · cs.CVROD-Dataset: A Pedestrian-Perspective Benchmark of Urban Obstacle Classes for YOLO-Based DetectionAbtin Zandi, Ariyan Azami, Bardia Sabagh Kermani, Parsa Abbasian +3
Two groups of pedestrians face avoidable risk on the city sidewalk: people living with visual impairment, who reach an obstacle before anything warns them, and people walking with divided attention, who register a hazard late or not at all. A vision system can warn them only about obstacles it has learned to recognize, so its training data limits the protection it offers. No suitable training set exists: public benchmarks are recorded from vehicles, aerial platforms, or fixed rigs, general-purpose detectors omit the street furniture that causes trips, and pedestrian-focused collections are con
benchmark - arxiv:2611.00676 · cs.CVSmooth bounding box for object detectionChao Zhou, Shenghua He, Fengchuang Xing, Han Yang +1
Deep learning-based object detection models are highly dependent on the quality of annotated training data. However, large-scale datasets often contain inevitable label noise, with bounding box localization errors being a prevalent and impactful form in object detection. While label noise handling techniques are well-studied for image classification, their adaptation to the more complex problem of bounding box errors remains underexplored. To address this, we propose a novel Smooth Bounding Box (SBB) strategy to mitigate the adverse effects of bounding box annotation inaccuracies. Our key insi
benchmark - arxiv:2611.00671 · cs.CVFrom White Canes to Edge AI: A Survey of Assistive Navigation Technologies for Blind and Low-Vision IndividualsAbtin Zandi, Arian Ashrafi, Sarvin Nami, Seyed Sina Negahban +7
Independent mobility is a foundational requirement for the autonomy of blind and low-vision (BLV) individuals, yet the assistive technologies that support it have historically evolved slowly. This survey systematically reviews 108 peer-reviewed studies published between 2008 and 2026, organizing the field into four technological eras: (i) a foundational era of passive mechanical aids, namely the white cane and the guide dog; (ii) a hardware era of electronic travel aids built on ultrasonic, infrared, and satellite positioning sensors that alert users to proximity but lack semantic understandin
benchmarkevaluation protocol - arxiv:2611.00658 · cs.CVSE-Mem: Semantic-Episodic Memory for Scene-Consistent Generative View SynthesisChangsong Lei, Yihan Li, Yu-Hui Wen, Huankun Sheng +4
Generative view synthesis aims to synthesize novel-view videos from reference observations along user-defined camera trajectories. Maintaining global scene consistency throughout this process remains challenging for camera-conditioned interactive video generators. Existing approaches typically select reference frames based solely on camera poses, overlooking whether the content remains semantically relevant or even visible. To address these limitations, we draw inspiration from semantic and episodic memory in human cognition, in which semantic memory captures generalized knowledge across exper
memoryepisodic memorysemantic memorybenchmark - arxiv:2611.00654 · cs.LGGateTrim: Adaptive-Rank Quantization for Large Language ModelsHongwei Ren, Jincheng Ren, Yuchuan Zhou, Jingchun Ling
We introduce GateTrim, an adaptive-rank post-training quantization method for large language models that learns matrix-specific ranks through gate optimization and threshold-based rank selection. GateTrim standardizes the low-rank factors of a quantized low-rank--residual representation, learns a separate gate vector for each target weight matrix, and uses a shared threshold to automatically determine its retained rank. The resulting ranks can differ across matrices, and the compact quantized representations are subsequently reconstructed. We evaluate GateTrim under W4A8 on Qwen3-4B/8B/14B, Ll
post-training - arxiv:2611.00653 · cs.LGRegularized Wasserstein Distance for Discrete Generative Modeling and Clustering AnalysisSenmao Ye, Fei Liu, Dawei Zhang, Yuqian Zheng +2
In this paper, we formulate clustering analysis as discrete generative modeling. Specifically, when the generator is restricted to a finite candidate space, the generated samples naturally emerge as the most representative prototypes of the underlying data distribution. To model such a discrete candidate space, we employ a Discrete Distribution Network (DDN). However, DDN suffers from unstable training due to its hard nearest-neighbor selection. To stabilize training and encourage geometric consistency, we replace DDN's l2 loss between images with regularized Wasserstein distance between the e
benchmark - arxiv:2611.00652 · cs.LGInference First, Training Underneath: Layer-Granular LoRA Co-Serving with LayerLoomZongpu Zhang, Yishu Yin, Xuehai Qian
Serving and fine-tuning a large language model on the same GPUs can turn the idle capacity between inference bursts into training progress, and low-rank adaptation (LoRA) makes this co-location cheap because both workloads read a single frozen copy of the base weights. The difficulty lies in the fine-tuning backward pass, which has no counterpart in inference. Current co-serving systems schedule backward work per serving iteration: each iteration carries a fixed backward budget, and requests that arrive in the meantime wait for it to finish, so fine-tuning dictates inference latency whenever l
memory - arxiv:2611.00650 · cs.LGHeirloom: Carrying a Phone's Personalization Across Base-Model GenerationsZhuomin Chen, Jingchao Ni, Qingkai Dong, Xu Zheng +4
On-device language models can learn a user's preferences through adapters trained on private data. However, a base-model upgrade may change the layers or vocabulary, leaving the personal adapter incompatible. Replacing it with a non-personal adapter loses much of the personalization, while retraining requires retaining the user's labeled history. We present Heirloom, a system that transfers a personal adapter to a new base model on the phone without retaining labels. Heirloom uses the old personal model as a teacher, collecting its predictions on stored unlabeled inputs, vendor-provided prompt
memory - arxiv:2611.00645 · cs.AIFastAgent: A Physics-Aware FastFluent-to-Fluent Workflow for Rapid CFD Screening and Industrial Steam Dewaxing DesignChen YanKai, Yu QiuHeng, Yuan Man, Zou Ting +5
Industrial computational fluid dynamics (CFD) becomes costly when complex geometries, long transients and coupled physics must be evaluated across large design spaces. We introduce FastAgent, a physics-aware campaign controller built around FastFluent, a lightweight numerical engine for rapid physical screening. FastFluent supplies stage-specific fields and diagnostics; FastAgent allocates physical models, numerical resolution and simulation horizons by the active quantities of interest (QoIs), while returned Fluent discrepancies update subsequent decisions. Across six public CFD benchmark cam
agentbenchmark - arxiv:2611.00643 · cs.LGSelecting LLM-Generated Hypotheses Raises Truth Beyond What an LLM Panel Expects: A Pre-registered Study on NAS-Bench-201 and HPOBenchShanglin Yang, Zhan Shi
LLM research-ideation systems aim for hypotheses that are both true and non-obvious, yet neither property can normally be checked. We build a testbed in which both can. Generators propose conditional interaction claims about two fully tabulated benchmarks, NAS-Bench-201 and HPOBench; every claim is verified exactly against the tables, and its non-obviousness is scored against a calibrated blind prior elicited from other LLMs. Excess truth measures how much more often a run's claims hold than this panel expects. Across a series of pre-registered stages with Claude and GPT generators, our main p
benchmark - arxiv:2611.00638 · cs.LGElectroChemLLM: A Multimodal LLM Pipeline for High-Fidelity Electrochemical InformaticsDikshya Mohanty, Manoj Praveen Nandigama, Kuldeepsinh Raj, Sourav Kanti Maiti +9
Electrochemistry is central to solving many modern-day challenges, including energy storage systems for a sustainable future, and its advancement relies heavily on the data from past findings. Existing automated pipelines are limited by text-only architectures, context fragmentation, and numerical hallucinations - collectively preventing accurate, large-scale curation of property databases. We introduce ElectrochemLLM, a multimodal information extraction pipeline that jointly identifies structured electrochemical records encompassing electrolyte compositions, electrode materials, cell configur
benchmark - arxiv:2611.00634 · cs.ROPREM: Prediction-Regulated Memory for Robot ManipulationXiao Huang, Ya Zhang, Weidi Xie
A robot's next action can depend on recent motion or earlier visual evidence absent from its current observation. A recurrent visual-memory architecture can retain this history in a visual state, but its usefulness depends on what the state represents and how it is updated with new observations. We introduce Prediction-Regulated Memory (PREM), a framework that uses action-conditioned future prediction to guide both memory learning and updating in vision-language-action policies. During training, predicting future visual features provides temporal supervision for memory learning. During executi
vision-language-actionmanipulationliberoaction-conditionedmemorymemory architecture - arxiv:2611.00633 · cs.AIEvidence-grounded large language models guide perovskite solar cell design and optimizationYanKai Chen, Zhi Wan, ZiHao Li, Tao Jing
Using large language models to interpret experimental evidence, design discriminating experiments and formulate well-supported mechanistic explanations remains a central challenge in materials research. Here we establish PSM-Bench, an 800-question benchmark developed by perovskite researchers to evaluate scientific reasoning in perovskite photovoltaics. We also introduce OpenPerov, an expert-informed language-model framework that learns from expert-calibrated evidence records reconstructed from 4,532 articles; these records link experimental strategies, controls and observations to mechanistic
benchmark - arxiv:2611.00631 · cs.AITopoQuant: Topological Framework to Characterize Behavioral Shifts in Quantized Large Language ModelsBaha Rababah, Shahzeb Qamar, Lorenz Sparrenberg, Rafet Sifa +3
Post-training quantization (PTQ) reduces the com putational cost of large language models (LLMs), but it can introduce internal structural and behavioral changes that are not captured by traditional metrics such as accuracy and per plexity. We present TopoQuant, a topological framework that uses topological data analysis (TDA) to quantify structural drift in self-attention weights and relate it to behavioral changes in quantized LLMs. Across four LLM families and multiple quantization levels, we show that topological drift remains stable at moderate compression levels but increases sharply bel
post-training - arxiv:2611.00627 · cs.AIAutonomous Analysis of Thermoelectric Materials Big Data by Multi-Agentic AI and Construction of a Knowledge Base and Specialized AIYusuke Hashimoto, Takaaki Tomai
Extracting knowledge from large databases of experimental materials data requires repeated analyses and reusable records of their choices and results. We developed Multi-Agentic AI (MAA), a multi-agent system in which AI agents with distinct roles analyze experimental thermoelectric data under predefined instructions and quality checks, saving analysis records traceable to their sources. In four experiments varying theme and cycle counts, we measured what accumulated in these records. In the largest experiment (100 analyses across 10 themes and 10 cycles), concept co-occurrences were concentra
ai agentmulti-agentagenticagent system - arxiv:2611.00624 · cs.LGXLA from the Trenches: Correctness on TPU Workloads at ScaleIgnacio Cano, Amir Saadat, Rishipal Singh Bhatia, Samuel Ginzburg +18
Accelerating modern foundational models has pushed domain-specific compilers toward increasingly complex, multi-tiered optimization pipelines. However, this heavy optimization focus introduces delicate tradeoffs between peak execution performance and correctness. Subtle codegen defects can trigger numerical degradation and memory or buffer corruption, invalidating multi-million-dollar training runs and degrading live serving quality without raising obvious compiler errors. In this paper, we present a battle-tested methodology for verifying and triaging Google's XLA:TPU compiler at hyperscale.
memoryagentai agent - arxiv:2611.00621 · cs.LGLog-Driven Semantic Tuning for Mixed Integer Programming SolversZhengzhong Ricky You, Kaizhao Sun, Zeping Min, Wotao Yin +1
Automated configuration of mixed-integer programming (MIP) solvers largely relies on black-box optimization. Each trial requires an expensive solver run and provides only a scalar outcome, so tuning can require many such evaluations. Human experts take a different approach. They inspect solver logs, identify performance bottlenecks, and adjust the parameters that address them. We propose Log-Driven Semantic Tuning (LDST), which uses a large language model (LLM) to automate this process. LDST extracts compact diagnostics from solver logs, uses the LLM to propose sparse edits to the current best
benchmark - arxiv:2611.00620 · cs.AIDiscovering Models for Experimental Quantum Systems with Agentic AIHannah C. Kleidermacher, Bennet Windt, Miguel Bello, James Zou +1
A central obstacle in realizing color center cavity-QED-based quantum technologies is inhomogeneity, or disorder, among individual emitters. As the number of emitters increases, so does the complexity of the experimental data, making it difficult to quantify the present physical mechanisms of disorder. Here we introduce an artificial intelligence-driven model discovery framework in which a large language model (LLM)-based agent, equipped with tools and a structured workflow, proposes candidate Hamiltonian and Lindbladian models, generates computational tools, and quantitatively evaluates compe
agentllm agentagentic - arxiv:2611.00619 · cs.AILM6-CastBench: A Geometry-Aware Simulation Dataset for Transient Thermal Prediction in LM6 Sand CastingSuhani Kolhatkar, Anand Khandekar
Transient thermal fields during solidification connect casting process conditions and cooling behaviour to the formation of defect-prone regions in aluminum cast ings. Repeatedly solving this thermal problem across geometries and process conditions is computationally expensive, motivating data-driven surrogate models. However, existing machine-learning studies on casting thermal behaviour rely on study-specific simulation databases that do not support reproducible geometry-level benchmarking. We introduce LM6-CastBench, an open simulation dataset for geometry-aware learning of transient therma
benchmark - arxiv:2611.00616 · cs.CLEnhancing Language Model Reasoning with Structured Multi-Level ModelingSiheng Xiong, Ali Payani, Faramarz Fekri
Inference-time scaling enhances a model's reasoning by extending its chain-of-thought (CoT). However, existing approaches typically rely on a single policy trained with outcome-reward reinforcement learning (RL), which often suffers from long-horizon plan failures where the implicit plan drifts from valid strategies, especially for small LMs with limited capacity. To address this, we propose Multi-Level Reasoning (MLR), which reformulates long-CoT generation as a two-level stochastic process. A high-level planner generates structured step descriptors specifying both the reasoning mode and the
benchmark - arxiv:2611.00615 · cs.ROLearning from an Imitable Teacher: Competition Solution for the MetaFine ChallengeYanze Zhang, Hao Liu, Haochang Hao, Zhixiang Shen +2
The MetaFine Open Competition (RoboFineMani workshop, CoRL 2026) evaluates fine-grained manipulation beyond binary success. Its five simulated tasks, from grasping a bottle by an instructed part to inserting letter-shaped pegs with 4 mm clearance, are scored on perception under lighting and camera changes, instruction understanding, and behavior. This report presents Team Friday's solution, which got second place with an overall score of 0.947, more than twice the $π_0$ baseline. Instead of fine-tuning a large vision-language-action (VLA) model, we train one 19M-parameter policy per task with
vision-language-actionmanipulationgrasp - arxiv:2611.00612 · cs.RORoz: A Taste Test for World Model LatentsKenneth Ge, Dhruv Sheth, Kunvar Thaman
State of the art latent dynamics world models achieve impressive performance for robot planning. But do they actually encode the physics of the world? In this paper, we show that popular world models today achieve strong results on tasks that reduce to image similarity but forget key details about objects in the scene like mass, deformation, friction, and bounciness. We achieve this by defining precisely what ought to be encoded in such a world model and introducing two new evaluation methods that can directly tease physical properties and easily scale to new ones. Our evaluation, which we cal
world modellatent dynamicsbenchmarkeval - arxiv:2611.00610 · cs.LGTrail: Accurate Local Smart Home Control Using Decision ModelsMaleeha Masood, Deepak Vasisht, Indranil Gupta
Natural-language driven smart home control remains challenging: cloud-hosted Large Language Models (LLMs) raise privacy concerns, while local Small Language Models (SLMs) can generate invalid actions or misinterpret requests. We introduce Trail (Typed Registry-Anchored Inference for Language commands in Smart Homes), a fully local framework that replaces open-ended command generation with hierarchical typed decisions. Trail exploits the natural structure of a home's action space, selecting a room, device, and operation from candidates derived from the runtime device registry. This ensures vali
memorymemory module - arxiv:2611.00608 · cs.CVWords Before Pixels: Selective Modality Routing for Vision-Language Model UnlearningLaura Yao, Haochen Zhang, Jinhao Duan, Sijia Liu +1
Vision-language model (VLM) unlearning is often treated as an objective-design problem: given multimodal forget data, the goal is to optimize a loss that removes undesirable behavior while preserving retained capabilities. We argue that this view overlooks an equally important data-centric question: how should each forget example be represented before optimization? Existing methods typically keep image-conditioned failures in their original image-text form, implicitly assuming that pixels are the right route for unlearning whenever pixels appear in the input. We challenge this assumption by co
benchmark - arxiv:2611.00605 · cs.AIConversational Interaction for LLM-Based Clinical Text-to-SQL: A User Study with Healthcare ProfessionalsJun-Min Lee, Suyoung Yoo, Gyubok Lee, Edward Choi +1
Conversational text-to-SQL interfaces could give healthcare professionals natural-language access to electronic health record (EHR) data, but evidence for their real-world usability remains limited: prior evaluations have relied largely on curated benchmarks with fully specified questions or on simulated users. We report a user study in which 14 physicians who regularly work with EHR data used both a single-turn and a conversational multi-turn text-to-SQL interface over the MIMIC-IV database in a within-subjects design; dialogue analyses additionally drew on interaction logs from 21 healthcare
benchmark - arxiv:2611.00602 · cs.AIEditing the Goal: Per-Token SAE Steering and Its Depth Limits in $π_{0.5}$Rocio Perales Valdes, Animesh Garg
Sparse autoencoders (SAEs) have recovered interpretable, causally relevant features in robot foundation models (RFMs) such as $π_{0.5}$ (arXiv:2603.19183, arXiv:2603.19233). SAE features on individual instruction tokens are less interpretable than timestep-pooled ones (arXiv:2603.19183), and whether they can redirect a robot's goal remains untested. We show that replacing the SAE features of a single goal word with those of an alternative goal word, leaving the instruction text unchanged, steers $π_{0.5}$-LIBERO into completing the alternative task. At layer 0, BatchTopK feature swaps redirect
pi0robot foundation modellibero - arxiv:2611.00596 · cs.AIX-World 1.6: Building Controllable Virtual Cities with World ModelsXPeng, :, Boyang Wang, Boyi Sun +32
Deploying generative world models for end-to-end closed-loop evaluation and policy learning requires precise scenario control, consistent observations, and sustained generation quality during extended interactions. This report presents \emph{X-World~1.6}, advancing X-World~1.0 toward a more reliable and versatile driving simulator. The upgraded base model offers finer-grained control over traffic semantics, including traffic lights and road markings, together with improved spatial consistency across cameras. Virtual City capabilities broaden scenario construction through a family of models con
world modelpost-training - arxiv:2611.00590 · cs.AIState-Aware Activation Recomputation with Replay Contracts and Selective PreservationAnnan Zhang, Shengjie Song, Wenshuai Yao, Lihao Zhang +1
Activation recomputation saves accelerator memory by replaying forward segments during the backward pass, but replay is only correct if it sees the state the original forward pass saw. Mutable buffers, aliased bytes and random-generator state can change in between, so re-executing unchanged code can differentiate a different computation. We make this state explicit. A versioned replay contract specifies what replay must restore, and a joint planner chooses, per segment, between activation retention, whole-buffer snapshots and selective-region preservation, charging each option its measured mem
memory - arxiv:2611.00586 · cs.LGDataset Mismatch Matters in Group Relative Policy Optimization for Reinforcement Learning from Verifiable RewardsZhen-Yu Zhang, Jiandong Zhang, Huaxiu Yao, Gang Niu +1
Recently, reinforcement learning from verifiable rewards (RLVR) is practically very important, where the leading approach is group relative policy optimization (GRPO) that typically faces a trade-off between rollout efficiency and downstream performance. To improve efficiency while preserving performance, existing selective rollout methods mainly exploit only source-side training dynamics to optimize the discrete distribution over source prompts. However, in practice, the source training dataset is rarely identically distributed to the target test dataset. When such a mismatch exists, due to l
benchmark - arxiv:2611.00580 · cs.LGEvolving Scientific Intelligence: From Fragmented Evidence to Distributed Experimental LoopsYuheng Ding
Scientific discovery advances when evidence from different investigations is brought into contact. Yet paper-centric communication leaves experimental results fragmented, delayed, or unreported, creating a "scientific memory wall" between growing computational capabilities and the evidence accessible to AI systems. We envision a different dynamic: in a network of intelligent laboratories, a failed or overlooked experiment in one domain could trigger a hypothesis in another, whose testing generates evidence that stimulates further investigations. Such chains of information collision and empiric
memoryevaluator - arxiv:2611.00562 · cs.LGTabAbax: A Tabular Foundation Model Technical ReportZhixiong Cheng, Xiaoyun Gao, Shiyuan Han, Jingjing Suo +1
Tabular foundation models (TFMs) enable in-context prediction on unseen tabular datasets without task-specific training. Mechanistic work has shown that increasing the depth of the ICL Transformer yields weak gains, leaving the main headroom for gains in the synthetic pretraining prior and column-row encoder upstream of the ICL backbone. We introduce TabAbax, a TFM built on TabICLv2 that focuses all modifications on these two upstream components. On the prior side, we further improve the quality and complexity of the synthetic pretraining prior. On the encoder side, a two-pass column-row stack
benchmark - arxiv:2611.00558 · cs.LGCan agents reliably design therapeutic antibodies?Asli Alpman, Nick Bhattacharya, Aaron Solomon, Hunter Elliott
Recent work has demonstrated agents driving protein design workflows, either through orchestration of specialized tools or direct manipulation of sequences. Yet the fidelity with which agents follow protein design instructions and the factors contributing to their effectiveness remain understudied. We examine agentic antibody design under two settings: when explicit molecular-editing instructions are provided and when agents receive only previously-scored sequences and are prompted to propose improvements. We call the first task-guided and the second black-box design. For the task-guided setti
manipulationagentic - arxiv:2611.00555 · cs.CLDistinguishing among reasoning levels of GPT 5.5 Pro via a grammar lens benchmarkEmma Graham, Richard Granger
Does the level of test-time processing effort affect a language model's computational capability? A formal grammar-lens benchmark method, derived from the theory of computation, generates novel data at specified levels, each successively more computationally powerful than the levels below. Using this data, we tested GPT-5.5, and GPT-5.5-Pro-Medium, -High, and -XHigh, showing incrementally increased acceptance of test data: 16, 40, 43, and 52\%, respectively, illustrating the benchmark's ability to discern fine-grained differences among models' added test-time reasoning levels. Additionally, as
benchmark - arxiv:2611.00554 · cs.LGFrom Weeks to Hours: Fast and Principled SFT Data Generation for Reasoning LLMsHongyi Henry Jin, Wenhan Yang, Meysam Ghaffari, Carlos Morato +1
Reasoning-oriented post-training enables large language models (LLMs) to solve complex tasks via multi-step inference. Supervised fine-tuning (SFT) on high-quality reasoning traces is a particularly efficient approach, but its effectiveness depends critically on careful data curation. Existing pipelines rely on a costly generate-then-filter paradigm: large pools of long reasoning traces are produced by multiple teacher models and subsequently filtered for difficulty and diversity using additional LLMs, often requiring weeks of computation. In this work, we propose SFTProbe, a fundamentally dif
post-training - arxiv:2611.00547 · cs.LGBitLLM: Efficient Low-Bit, Conversion-Hiding LLM Decoding on Mobile GPUsHangyu Zheng, Zhihao Shu, Kunxiong Zhu, Minghai Qin +3
The rapid scaling of large language models (LLMs) has unlocked remarkable capabilities across various domains. However, deploying LLMs on mobile devices remains challenging because of their high computational and memory requirements. Low-bit quantization (e.g., 4-bit weights, 8-bit activations) can reduce model size while preserving accuracy, but its practical efficacy on mobile GPUs has been limited so far. In this paper, we identify that the performance gap stems primarily from frequent data type conversions between quantized and floating-point values, and dynamic and inefficient memory acce
memorylong-context - arxiv:2611.00546 · cs.AILow-Margin Branch Points in Deterministic LLM DeliberatioHajime Shimao, Warut Khern-am-nua
Deterministic multi-agent LLM deliberation can exactly re- play a fixed transcript, yet exact replay alone does not test local resistance to controlled internal interventions. Because each response becomes context for later agents, a changed greedy token can redirect a transcript-mediated execution. We present a white-box audit of this local failure mode in a five-agent committee under greedy decoding. We prospectively select low- and high-margin token positions, require exact no-op re- play, inject norm-relative residual displacements, and measure first transcript divergence within ten genera
multi-agentagent system - arxiv:2611.00544 · cs.AISkill++: Towards Lifelong Agents via Routing-Aware Skill RefinementHanrong Zhang, Minghao Liang, Jiawei Wu, Yankai Chen +9
Large language model (LLM) agents extend their capabilities with skills, each with a name, a description, and a body. Under progressive disclosure, an LLM router selects skills by name and description alone and loads only the selected bodies. Skill libraries grow continually as users and agents add new skills. Because the router relies on this metadata, a similar new skill can displace an earlier one, breaking previously correct routing even though the earlier skill and the router are unchanged. This routing interference is aggravated for public skills, whose descriptions may not match their b
lifelong agentagentbenchmark - arxiv:2611.00539 · cs.ROFaster Imitation by Re-timing Demonstrations Within the Controller LimitZongtan Li, Tianming Liu, Ye Zhao
Imitation policies inherit the operator's pace, and re-timing the demonstrations to run faster works only until success collapses at two to three times demonstration speed. Recent methods attribute that collapse to distribution shift and answer it with finer rules for which segments may be sped up; we show it has a mechanical cause that segment selection cannot address. A low-level controller realises only a bounded displacement per control period, so merging consecutive commands produces sums outside that bound: training pairs that promise a transition the plant cannot deliver. The share of s
humanoid - arxiv:2611.00536 · cs.CLDual-Axis Policy Entailment: A Benchmark for Policy-Change ReasoningCharles Chang, Yuen Yuen Ang, Bolun Sun
Whether a policy shifts domains and, by extension, continues, broadens, narrows, or reverses previous policy is of paramount importance to the study of public policy. We propose Dual-Axis Policy Entailment (DAPE), which evaluates an ordered pair of policy directives on both dimensions at once: a binary judgment determines whether the two directives govern the same policy domain, and a five-way judgment determines whether the later directive continues, broadens, narrows, or reverses the earlier policy, or governs a different domain altogether. Read together, the two judgments form one five-cate
benchmark - arxiv:2611.00527 · cs.RO3D Neural Rendering for Automated Ultrasonic Laser ScansThomas Pierson, Matthew Mitchell, Kira Perna, Anand Iyer +2
Laser doppler vibrometer (LDV) scans can be used to detect defects and cracks in manufactured parts before they are deployed. This work presents a method for reconstructing the steady-state surface wavefield of a three-dimensional part by combining automated robotic manipulation with neural rendering. A six-degree-of-freedom (6-DOF) robotic manipulator (UFactory xArm6) with an embedded piezoelectric transducer presented an acoustically excited test article to a stationary LDV at a sequence of known orientations, yielding a set of complex-valued 2D scans with poses derived analytically from the
manipulationmanipulator - arxiv:2611.00524 · cs.ROReflex: Fast Robot Policy Serving by Reusing and Correcting Past ActionsZiyang Lin, Zixuan Sun, Rami Nasser, Jianli Jin +3
Modern robot policies often perform expensive inference to generate action chunks, making serving a latency bottleneck for real-time control and a throughput bottleneck for multi-robot serving. Building on the success of caching in image, video, and audio generation, recent robot-policy advances have reported promising speedups by reusing intermediate denoising states to avoid redundant computation. Surprisingly, we find that such reuse often offers little advantage over simply running fewer denoising steps from fresh noise, without any cache. Moreover, accelerating denoising leaves the expens
action headrobot policy - arxiv:2611.00522 · cs.ROBehavioral Steering of Vision-Language-Action Models in the Low-Parameter RegimeJames Tcheng, Feiyu Gavin Zhu
Reinforcement learning (RL) is increasingly used to adapt pretrained vision-language-action models (VLAs). We investigate the problem of behavioral steering: adapting a VLA's execution style to preferences such as carrying objects higher. To investigate how little trainable capacity such steering requires and how much pretrained capability can be retained, we freeze an OpenVLA-OFT 7B policy and optimize TinyLoRA adapters with 8 to 384 trainable parameters using either policy-gradient RL (GRPO) or evolutionary search (CMA-ES). With just 256 trainable parameters, we raise carry height by 120 mm
vision-language-actionvlaopenvla - arxiv:2611.00518 · cs.AIAgentic Context Management for Multi-Turn Human-AI ConversationsMatthew Ho, Chenglong Wang, Baolin Peng, Nebojsa Jojic +4
LLMs degrade in multi-turn conversation partly because of context pollution: the model's own earlier outputs stay in context and anchor later turns, so an early misreading of the user's intent can persist after the user corrects it. Assistant omission (AO), dropping every assistant message, removes this pollution but fails when interaction is referential: later turns refer back to drafts, decisions, and tool-call results that exist only in assistant turns. We propose Agentic Conversation Context Curation (AC3), a training-free method in which a separate analyzer (1) consolidates the task speci
agentictool use - arxiv:2611.00515 · cs.AIAStockArena: Beyond Returns, Can We Entrust Real Capital to the Decisions of Financial Agents?Zeyu Liu, Sing Kwong Lai, Xuanzhi Feng, HaoYang Shang +3
Large language models (LLMs) are increasingly used in dynamic financial trading, yet real-time trading benchmarks mainly evaluate agents by cumulative returns and risk-adjusted performance. Such outcome metrics cannot show whether trades are faithfully reported, supported by available evidence, and consistent with account states. We introduce AStockArena, a real-time decision-auditing benchmark for continuous trading in China's A-share market. Under shared market conditions and execution constraints, we evaluate 13 open- and closed-source models over 3,198 sequential decisions. AStockArena rec
benchmark - arxiv:2611.00513 · cs.ROWorldSketch: Sketching World Models for Zero-Shot Mobile ManipulationJiaming Wang
Most agentic robotic systems use vision-language models (VLMs) either to orchestrate predefined skills or to directly generate actions, constraints, and control programs. The former is limited by the available skill library, while the latter requires VLMs to reason about precise geometry and contact. In this paper, we propose WorldSketch, a new paradigm for agentic robotics in which a VLM constructs a world model rather than generating robot actions. The VLM acts as an agent that actively queries perception tools to build and visually validate a minimal kinematic world model, capturing only ta
manipulationliberoworld modelagentagenticself-improving - arxiv:2611.00508 · cs.AIDo Social Influence Pressures Move LLM Agents? Evidence from Delegated Audit JudgmentRichard Zhe Wang
Large language model (LLM) agents are rapidly being integrated into complex professional workflows such as auditing and entrusted with judgments under human review. Human auditors are known to yield to social influence pressures from superiors, peers and clients, and this paper asks whether LLM agents yield to the same pressures and, if so, how they respond. In an experiment on twelve model settings, an LLM agent acting as a senior associate in audit analytical review forms an estimate and is then challenged without new evidence by colleagues inside the firm, by a reviewer from another firm an
agentllm agent - arxiv:2611.00507 · cs.CVSelecting or Solving? What Best-of-N Verification Delivers in Video ReasoningMartin Q. Ma, Yuxiao Qu, Ruslan Salakhutdinov, Louis-Philippe Morency +2
Answering a complex question about a long video is a grounding problem: the answer turns on a few seconds of evidence, and vision-language models remain well below human accuracy at finding them. The standard way to spend extra test-time compute on this kind of video reasoning is best-of-N selection: sample several chain-of-thought answers from a fixed model, then let a second model re-watch the video and pick one. Such pipelines are reported by a single headline number, their margin over a majority vote. That number conflates two effects with opposite deployment consequences: the benefit the
benchmark - arxiv:2611.00505 · cs.AITo Jev or Not to Jev: When Can System-1 Decision Models Replace LLM Judges?Shreyas Subramanian
Decision models such as TypeSafe's Jev answer typed questions about a text (yes/no, multiple choice, or an ordinal score) with calibrated probabilities in a single forward pass and without generating text, which makes them an inexpensive candidate for LLM-as-a-judge workloads. We evaluate 14 System-1 decision models, selected with a pre-stated protocol, against four generative judges (Claude Haiku 5.5, GPT-5.6 with and without reasoning, and Claude Opus 5.5 with thinking) on eleven established judge benchmarks. Five of the 14 are statistically indistinguishable from GPT-5.6 without reasoning o
benchmarkevaluatorleaderboard - arxiv:2611.00504 · cs.AIRobustness Starts with the Objective: Designing and Certifying Robust LLM-Generated HeuristicsPantea Karimi, Roozbeh Bostandoost, Dany Rouhana, Siva Kesava Reddy Kakarla +4
LLMs are increasingly used to design heuristics for computer systems. However, most synthesis frameworks evaluate candidates by their average performance, which can mask severe regressions on specific workloads. Even adding adversarial training inputs doesn't eliminate these regressions, and percentile objectives (e.g., optimizing for the 95th percentile) overlook how severe regressions can be beyond the selected threshold. We present TAILOR, a framework that guides LLM synthesis using the Conditional Value at Risk (CVaR) of the cost difference between each candidate and an existing baseline o
benchmark - arxiv:2611.00499 · cs.ROKnowing Where to Look: Interactive Object Search via Spatial-Semantic FusionAditya Agarwal, Lucy Cai, Kelly Lu, Travis Manderson +3
When searching for an object, humans naturally combine semantic intuition - reasoning about where the object is likely to be stored - with geometric awareness of what parts of the space are visible, occluded, or unexplored. Effective robot object search requires the same joint reasoning, yet existing methods rely on only one: geometric methods maintain principled spatial beliefs but lack deep semantic understanding, while language-model based methods leverage commonsense knowledge but maintain no explicit belief over space. We argue that jointly reasoning over semantic and geometric evidence i
humanoid - arxiv:2611.00496 · cs.AISwap, Replicate, Read the Mass: A Protocol for Auditing Demographic Bias in System One Decision ModelsAlejandro Mozo Quesada, Sergio Gálvez Rojas, Francisco López Valverde
System One decision models replace text generation with typed, probability-scored answers and are entering production. Answers execute without a person in the loop, and no multi-model audit exists on established bias benchmarks. We provide that audit: three benchmarks: BBQ (all eleven categories), CrowS-Pairs, and SilverSpoon, adapted to the typed interface, twelve models plus a no-thinking ablation, and 1,638,000 decisions. Every order-sensitive item runs twice with options swapped; native typed models also swap labels; ten configurations repeat under extra seeds and batch 16. We find widespr
benchmark - arxiv:2611.00492 · cs.AISpotting the Artificial: Dialogic Humor Improvisation in Humans and LLMsLizeth Palacios Patiño, Ximena Gutierrez-Vasques, Daniel Rodriguez
We propose a methodology for investigating the differences between natural and synthetic language in a challenging setting: humorous improvisational dialogue. Our experimental setting compares human-produced comedy improvisations with the dialogues generated by LLMs in a multi-agent environment that we designed. The analysis combines automated linguistic measures with manual annotation grounded in Systemic Functional Linguistics (SFL). The results show that, although LLMs can create coherent and creative scenes, they often fail to reproduce key characteristics of spontaneous spoken interaction
multi-agent - arxiv:2611.00489 · cs.AIThreat Modeling Intimate Partner Abuse in Agentic SystemsNatalie Grace Brigham, Eugene Bagdasarian, Franziska Roesner, Tadayoshi Kohno
Agentic systems are AI-powered technologies capable of autonomous action within digital environments, such as sending emails or making reservations. They are deeply connected to external tools as well as users' data and resources, making them vulnerable to attacks which exploit their access and capabilities. In this work, we apply the technology-facilitated abuse threat model, in the context of intimate partner violence and surveillance, to these systems, revealing threats that require deliberate investigation and mitigation. We categorize attacks based on system ownership: (1) Abuser-as-owner
agentic - arxiv:2611.00482 · cs.ROA Straight Line Is Not a Camera Pose: Auditing Short-Clip Pose BenchmarksXinyi Shi
The absolute trajectory error (ATE) is the standard measure of camera-motion accuracy, from SLAM to feed-forward 3D reconstruction. On the common Sintel protocol, a constant-velocity straight line using no images has a lower mean ATE than 10 of 11 recent models re-run with official code. Two properties of ATE's alignment explain this. First, it maps any constant-velocity prediction onto the best-fitting line through the true path, so the line scores the path's regression residual. A residual signal-to-noise ratio (SNR) criterion predicts when moving a prediction towards its constant-velocity f
benchmark - arxiv:2611.00478 · cs.CVSpatial Uncertainty Analytics for Budget-Limited Inspection of Hyperspectral ScenesRitesh Sharma, Yuichi Motai
Land-cover maps classified from wide-area hyperspectral surveys are least reliable along class boundaries and in ambiguous regions, so their errors are spatially structured. We study how this structure can direct a limited verification budget: when a low-altitude UAV or a ground crew can visit only a handful of sites, where it should go depends on how classifier uncertainty is arranged across the scene and how a fixed sensor footprint tiles that arrangement. We estimate a per-pixel uncertainty surface from the disagreement of an ensemble of spectral classifiers and select sites by greedy, fiel
benchmark - arxiv:2611.00474 · cs.AIFew Trials, Many Metrics: Beta Mixtures for Low-Budget AI EvaluationGustavo Tironi, Felipe Maia Polo, Diego Mesquita
Evaluating a stochastic AI system often requires many repeated samples per instance, especially for metrics defined over long horizons. Recent work has shown that modeling instance-level success probabilities with a Beta distribution recovers Pass@$k$ from far fewer samples than direct estimation. We show that this framework supports a broader range of evaluation targets. We fit a mixture of Beta distributions by marginal maximum likelihood, with the number of components chosen by BIC. From that one fit, we compute Pass$^k$ and AUC@$k$ at any horizon in closed form. We also derive delta-method
benchmark - arxiv:2611.00464 · cs.AICompression Technique Interactions in Small Language Models: A Factorial Ablation StudyBhushit Nema, Siddharth Krishnamurthy, Anjali Kadiyala
Deploying small language models on resource-constrained hardware requires compression, yet practitioners lack systematic guidance on how quantization, pruning, and post-compression recovery interact when combined. We present a factorial ablation study evaluating 72 compression configurations across four SLM architectures (2.7B-7B parameters), extended with a 48-configuration ordering analysis, benchmark accuracy validation on 30 compressed models, and robustness checks including multi-seed validation and full-corpus perplexity evaluation. INT8 quantization is effectively lossless (<0.5% accura
memorybenchmark - arxiv:2611.00463 · cs.AIExo: An Architecture for Self-Improving Systems at RuntimeAlexander Krentsel, Sylvia Ratnasamy, Ion Stoica, Martin Casado +1
Agent harnesses govern how language models act in an environment; designing their policies by hand is difficult, while automated techniques require representative datasets and costly executions. We present Exo, an agent runtime architecture for safe learning during ongoing work that collapses the optimization loop into task execution, allowing agents to revise both their harness and their strategies for further improvement. A minimal protected mechanism layer preserves execution history and governs policy activation while the harness evolves through inspectable code changes. On long-running ob
agentself-improving - arxiv:2611.00460 · cs.AIFrom Small to Large: Assessing Medical Reasoning, Benchmark Contamination, and Model Size in Language ModelsNavapat Nananukul, Mayank Kejriwal
Large language models (LLMs) now answer medical exam questions at or above expert level, but it is unclear whether such scores reflect clinical competence, or are instead affected by the format of the questions being asked (e.g., multiple-choice versus open-ended), prior exposure to the benchmarks, and the difficulty of the questions. These benchmarks are also public and widely used online, which may be resulting in LLM memorization. Models may perform better on questions that they have seen before. It is also not evident whether smaller models, which are more locally deployable in HIPAA-compl
benchmark - arxiv:2611.00458 · cs.CVPRISM-Net: Dynamic Routing Network with VLM-Guided Disentangled Attribute Distillation for Explainable Fine-Grained Product RetrievalChung-Ming Lo, Che-Wei Chang
Fine-grained visual product retrieval in large-scale e-commerce multimedia systems demands discriminative representations capable of simultaneously resolving macro-level category semantics and micro-level visual attributes such as structural contour (line) and chromatic appearance (color). However, conventional backbones including convolutional neural networks (CNNs) and vision Transformers (ViTs) inherently suffer from feature entanglement and shortcut learning. To learn the micro-level visual attributes without costly manual attribute annotations, we propose phase-routed interpretable semant
benchmark - arxiv:2611.00455 · cs.ROTacMind: Benchmarking and Modeling Tactile Interaction States for Robotic ManipulationJiaxin Shi, Zibin Wu, Yichi Zhang, Jianghua Duan +2
Tactile sensing provides robots with direct observations of contact physics, local geometry, and interaction dynamics, making it a vital source of feedback for contact-rich manipulation. However, existing tactile manipulation studies are often developed and evaluated within specific task settings, making it difficult to systematically compare how tactile information contributes under different interaction conditions. In this work, we develop a benchmark in simulation comprising 128 tasks organized around three complementary evaluation regimes: Precise, Recognition, and Dynamic. The benchmark f
manipulationtactileworld modelaction-conditionedbenchmark - arxiv:2611.00452 · cs.LGCompositional Self-Improvement: Promises and PitfallsChangho Shin, Daiwei Chen, Brenden Lake, Frederic Sala +1
Although self-improvement is becoming increasingly important as tasks become more complex, models struggle to generate reliable supervision for harder, out-of-distribution instances. One way to address this challenge is to exploit compositional structure: models often solve simpler subproblems more reliably than full problems, so their subproblem predictions can be composed into supervision for harder ones. We systematically study when this kind of composed supervision supports self-improvement and when it fails. Our controlled experiments identify three failure modes: unreliable subproblem pr
self-improvement - arxiv:2611.00450 · cs.LGFrom Lookups to Lessons: Turning Retrieved Knowledge into Parametric MemoryChangjiang Han, Bowei He, Ye Yuan, Weixu Zhang +3
Retrieval-augmented generation (RAG) helps large language models (LLMs) answer factual questions beyond their training data, but it adds storage and inference costs. Distilling previously retrieved information into parametric memories enables reuse without online retrieval. However, training the memory to mimic a retriever rewards accurate independent predictions, which do not necessarily improve the final output quality when combined with the LLM's existing predictions. To address this, we introduce ResMem, in which the LLM reads retrieved text offline to generate teacher targets. This design
memoryretrieval-augmented - arxiv:2611.00449 · cs.LGSpBench: An Agent-Native Framework for Cross-Engine Lakehouse BenchmarkingLyndon Shi, Nic Crouch, Yi Hu, Zihao Ye +6
SpBench is an agent-native framework for cross-engine lakehouse benchmarking, in production use for benchmarking Snowpark, Snowpark Connect for Apache Spark, and other execution environments across standardized suites -- such as TPCx-BB and TPC-DI -- and custom benchmarks distilled from production workload patterns. It runs continuously across a broad portfolio of workloads, guiding the performance roadmap and detecting regressions for production-scale data engines. Benchmarking has historically been an expensive investment requiring continuous cross-functional collaboration and domain experti
llm agentbenchmark - arxiv:2611.00447 · cs.AIQuantiles, Not Points: Streaming Quantile Matching for Ordinal-Aware LLM-as-a-Judge CalibrationIlija Subasic, Andrew Rabinovich, Zhao Chen
LLM-as-a-judge scores are often distributionally shifted from human ratings on any given evaluation task, and post-hoc calibration (mapping LLM to human scores in a small labeled anchor set) has become a popular method to correct for this shift. In this paper we derive the calibration from two positions: that LLMaJ scores are ordinal ratings drawn from a distribution rather than isolated points. We present streaming quantile matching (SQM), a version of a classical method not previously bench marked for LLMaJ calibration, which is built on the two positions by computing the optimal-transport m
evaluation frameworkjudge model - arxiv:2611.00444 · cs.LGCLiS: Plug-and-Play Attention Acceleration for Video Diffusion on TPUs via Certified StabilizationElisa Tsai, Shao-yu Huang, Daniel Mandragona, Wei-Yu Lin +5
High-resolution video diffusion relies on attention over long token sequences, making attention efficiency critical to generation latency. Our analysis identifies repeated maximum reductions and accumulator rescaling in online softmax as a bottleneck in long-context TPU attention. We propose CLiS (Certified Loop-invariant Stabilization), a training-free, plug-and-play optimization for video diffusion inference. CLiS fixes per-row stabilizers before the KV loop and certifies numerical safety per head, falling back to online softmax when certification fails. It integrates into existing fused-att
long-context - arxiv:2611.00442 · cs.AIFidelity Collapses with Table Size, Not Turn Count: Delivery Architecture in a Deployed Tool-Using AgentTung Thanh Hoang
When a multi-turn agent delivers a tool's structured records into the model's context and re-emits them each turn, every turn can corrupt data the tool computed correctly. We measure this in a production award-travel agent: two architectures over identical frozen payloads, lineage-matched prompts and sampling, differing only in delivery architecture. SYNTHESIS feeds full record sets into the planner's context and answers follow-ups in-head; BYPASS, the deployed system, keeps records in session state, translates follow-ups into arguments for a deterministic filter, and delivers rows over a type
agent - arxiv:2611.00438 · cs.LGStatistical Inference under Performativity: From Policy Feedback to Self-Evolving SystemsTianwei Gao, Garrett G. Wen, Huiying Zhong, Lihua Lei +2
Performativity arises when prediction-informed decisions change the data-generating process, creating feedback between deployment and learning. This phenomenon occurs in policy-making and in self-evolving AI systems that shape their own future training data. Statistical inference is challenging because estimation errors in earlier predictors affect subsequent data distributions and propagate into later updates. We develop an end-to-end framework for statistical inference under performativity. For repeated risk minimization and repeated gradient descent, we establish central limit theorems at f
self-evolving - arxiv:2611.00432 · cs.AIOnline Fair Division of Chores with Upgrade RecourseYichen Huang, Nicholas Teh
We study the online fair division of indivisible chores among $n$ agents under upgrade recourse, where earlier assignments may be revised but each chore previously assigned to an agent must be replaced by a distinct chore that she finds no more costly. We show that a simple adaptation of round robin maintains envy-freeness up to one chore (EF1) together with a $(2-1/n)$-approximation of the maximin share (MMS), and that no deterministic algorithm without predictions can do better, even when all agents rank the chores in the same order. This stands in contrast to the allocation of goods, where
agent - arxiv:2611.00430 · cs.CVReal-VAS: a Real-World Video Amodal Segmentation DatasetFinlay G. C. Hudson, William A. P. Smith
Amodal video object segmentation is fundamentally limited by the absence of datasets that combine real-world complexity with precise ground-truth annotations. To address this, we present Real Video Amodal Segmentation (Real-VAS), a new large-scale, zero-shot evaluation dataset built from composited real-image video. We introduce a novel data generation pipeline that enables the creation of programmatically consistent amodal ground truth for real-world footage, in which the object mask, occlusion boundary, and composited image are all derived from a single unoccluded source, without relying on
benchmark - arxiv:2611.00427 · cs.LGSpoofing the Latent State: Imperceptible Goal-Image Attacks Derail Planning in LeWorldModelKunal Bhosikar, Nicole Meng, Charu Sharma, Yingjie Lao
Joint-embedding predictive architectures (JEPAs) have recently emerged as general-purpose visual world models for control. A prominent example is LeWorldModel (LeWM), which trains end-to-end from pixels using an isotropic Gaussian anti-collapse regularizer (SIGReg) and plans by minimizing the squared Euclidean distance between predicted and goal embeddings. In this work, we investigate the test-time robustness of this latent planning interface. Analyzing the learned metric on Push-T, we first find that task-relevant object state heavily dominates latent distance: shifting the block by a single
world modelagentbenchmark - arxiv:2611.00420 · cs.CLCircuit Oracle: Automating Attribution Graph Analysis via Natural-Language QueriesShariar Kabir, Hong Kiat Tan, Swastik Agrawal, Sai V R Chereddy +1
Attribution graphs, an emerging tool in mechanistic interpretability, use transcoders to decompose language model computations into sparse interpretable features connected by causal edges. However, turning a graph into a safety-relevant insight requires hours of manual analysis by experts. We introduce Circuit Oracle, an off-the-shelf LLM equipped with a multi-agent harness and task-specific authored skills, which answers natural-language questions about a target model by reading the attribution graph through tool calls. We evaluate it on three safety-relevant proxy tasks: detecting spurious f
multi-agent - arxiv:2611.00408 · cs.LGTileWeave: Heterogeneous Tiling for Sparse Prefill on Mobile SoCsWentao Ni, Hong Cai, Mingu Lee, Chihao Yu +6
Smartphone applications increasingly require large language models to process long contexts, whereas high prefill latency and memory demands constrain local inference. Block-sparse attention reduces computation and memory traffic, but its irregular selections conflict with mobile NPUs' preference for large, regular tiles. Efficient execution must exploit cross-query reuse while avoiding redundant computation and preserving selections needed for accuracy. In this paper, we present TileWeave, which partitions sparse attention by this cross-query reuse: highly shared key blocks run in coarse NPU
memorylong context - arxiv:2611.00407 · cs.LGHacking or Just Failing? A More Capable LLM Monitor Better Separates Reward Hacking from Honest FailureKaushik Kezendirane
An agent's visible score can exceed its real progress because it gamed the score or because it simply made mistakes, so a useful reward-hacking monitor must tell the two apart. We built a sandboxed task in which an agent can game a completion score, labelled every episode by replaying it and repairing the final state counterfactually, and ran five pre-registered studies. On 60 episodes from a Llama 3.1 8B agent, small local monitors (Llama 3B and 8B) did not separate exploit attempts from honest failures with the same score gap: prompt changes traded missed exploits for false alarms, and the p
manipulationagent - arxiv:2611.00406 · cs.CVTAP-Out: Tracking Any Point in 360° via OutpaintingFinlay G. C. Hudson, James A. D. Gardner, William A. P. Smith
Tracking points that leave the camera's field of view (FoV) requires reasoning about unobserved scene dynamics, a fundamentally different problem from tracking visible correspondences. We present TAP-Out, an offline method that makes this out-of-view (OOV) reasoning explicit by leveraging generative video outpainting as a world-model prior. Given a narrow-FoV input video, TAP-Out outpaints a complete 360° equirectangular video, locates target objects within the panoramic scene via video object segmentation, and synthesises object-centred perspective crops on which standard in-frame trackers op
benchmark - arxiv:2611.00403 · cs.LGMXCode: Vector Quantization for MX-Native Inference at ScaleGaurav Kolhe, Cansu Demirkiran, Shubhi Tyagi, Sulaiman Sadiq +7
Memory capacity and bandwidth are major bottlenecks in LLM inference, motivating weight compression beyond conventional 4-bit quantization. Vector quantization (VQ) enables approximately 2-bit weight storage but often relies on codebooks incompatible with native FP4 computation. We introduce MXCode (MX-Native Codebook Quantization), a framework bridging ultra-low-bit VQ compression with MicroXcaling (MX)-native FP4 computation. MXCode proposes three strategies: PostScale, which quantizes learned codebooks to MXFP4; PreScale, which extracts MX scales before codebook learning; and PreScale+, whi
memory - arxiv:2611.00401 · cs.LGVESPER: Online Learning of Free Energies with a Multi-Tempered Variational ObjectiveJay L. Kaplan, Gustavo R. Pérez-Lemus, Mengjian Hua, Juan J. de Pablo +1
Computing molecular free-energy landscapes over collective variables (CVs) is central to understanding molecular mechanisms, but poses a difficult learning problem without training data available a priori. Samples of the marginal CV distribution must be generated by molecular dynamics (MD), whose trajectories can remain trapped in metastable states and rarely cross the free-energy barriers governing conformational changes or chemical reactions. We introduce VESPER, which couples the generation of these samples to neural-network-driven free-energy learning: MD supplies configurations for learni
online learning - arxiv:2611.00394 · cs.ROCalVLA: Shaping Native Confidence in Vision-Language-Action Policies with Outcome SupervisionBoniface Chukwuemeka Godwin
Can a vision-language-action (VLA) policy signal likely task failure from its own action probabilities? We introduce CalVLA, which shapes native action-token confidence using successful and failed rollouts and an Action-Confidence Alignment (ACal) squared margin-ranking loss. Low-rank adapters update OpenVLA-7B without an auxiliary failure detector. Observations are paired within tasks at similar normalized elapsed times, not necessarily equivalent physical progress. On a held-out test split from 500 LIBERO-10 configurations, adapted confidence achieves a joint failure-discrimination AUROC of
vision-language-actionopenvlalibero - arxiv:2611.00387 · cs.LGGeometric and algebraic inference in humans and neural networksJacob Russin, Joonhwa Kim, Ellie Pavlick, Michael J. Frank
Humans possess a remarkable ability to draw inferences from limited data. While some inferences are made intuitively from past experience, others require online reasoning based on novel information given in context. Here, we formalize these strategies, which we term geometric and algebraic inference, and study their computational tradeoffs in humans and neural networks. The geometric strategy gradually shapes representational geometry to reflect underlying relational structure, making subsequent inference cheap. The algebraic strategy flexibly attends to and reasons over novel relations given
manipulation - arxiv:2611.00386 · cs.AIDependence-Aware Risk Assessment and Optimal Design of AI-Agent Safeguard SystemsYijia Li, Maochao Xu, Peng Zhao
Large language model (LLM) agents increasingly rely on safeguards to prevent unauthorized actions while preserving legitimate service. However, existing risk assessments often focus on individual safeguard performance while overlooking dependence among safeguard decisions. This paper develops a dependence-aware framework for risk assessment and optimal design of multi-component AI safeguard systems, jointly selecting configurations and operating policies for sequential layered screening and selective verification. We establish stochastic comparisons of layered-screening error probabilities and
agentbenchmark - arxiv:2611.00384 · cs.AIAgent Stack for ML Training Lifecycle: Generation, Validation, and OptimizationJaideep Ray, Daniel Tang, Zejun Lin, Haoyue Tang +20
Production model development requires repeated architecture design, platform integration, debugging, training, and evaluation. We present an agent stack deployed at LinkedIn with two independently usable agent loops: AutoTorch for model generation and migration, and AutoResearch for model optimization. AutoTorch generates PyTorch models and training pipelines for TensorFlow migration, XGBoost replacement, and new-model development. Executable validation hooks check platform API compliance, trainability, and conversion fidelity on sampled data, guiding iterative repair before engineer review an
agentpost-training - arxiv:2611.00383 · cs.CVLF-VideoQA: Evaluating Long-Video Understanding with Long-Form Theory-of-Mind QuestionsShaden Shaar, Bradon Thymes, Majd Aldaye, Claire Cardie +1
The hardest part of understanding a long video is not pinpointing any one moment but synthesizing many, e.g., answering a question that requires tracking what develops across hours of content and articulating how the pieces fit together. To date, most long-video QA datasets require only a short answer or a multiple-choice selection, oversimplifying the long-video understanding task to retrieval. We introduce LF-VideoQA, a dataset for long-form, open-ended question answering over long narrative videos built on theory-of-mind questions: a character's belief depends on everything they have seen o
benchmark - arxiv:2611.00367 · cs.ROWhen appearance and motion disagree: evaluating class--motion consistency against disguised uncrewed surface vesselsJon Reifschneider
Defense systems for maritime assets must distinguish approaching benign craft from uncrewed surface vessels (USVs) with sufficient warning to act. Shore cameras using computer vision can separate hostile vessels from benign traffic, but these systems may be attacked using an adversarial hull patch or physical disguise. We evaluate adversarial hull patches and find they suppress detection in up to 29% of attempts but never cause misclassification as benign. Physical disguise, however, can bypass both a camera check and a radar presence check. To evaluate defenses, we develop a hybrid replay pro
evaluation protocol - arxiv:2611.00359 · cs.CLHumanized but Still True? Auditing Facts and Detection in AI-Text HumanizersZiyuan Wang, Genliang Zhu, Xiaobo Chen
AI "humanizers" rewrite machine-generated text so that detectors classify it as human-written. Public leaderboards score evasion and embedding similarity, but not whether checkable details survive. We check whether numbers, dates, person and organization names, and quoted spans survive each rewrite, using a rule-based checker validated by two AI annotators (precision 0.84-1.00, recall 0.93-0.95; no human annotation). In the October 2026 cycle of a public humanizer leaderboard, all 11 commercial tools changed or dropped at least one of these units in 35-65% of the documents that contained one.
leaderboard - arxiv:2611.00351 · cs.CVSIEVE: Preserving Cross-Image Evidence via Depth-Aware Token CompressionNitish Shukla, Surgan Jandial, Nhat-Tan Bui, Arun Ross
Multi-image vision-language models (VLMs) incur substantial computation when processing long visual sequences. Compressing these sequences must preserve complementary evidence across images and keep tokens from different images separate. Global selection can underrepresent question-relevant images, and pruning in early decoder layers risks removing relevant content before sufficient multimodal processing, yet existing methods spend most of the budget on these early layers. We propose SIEVE, a training-free, single-pass framework that jointly allocates visual tokens across images and decoder de
benchmark - arxiv:2611.00350 · cs.LGOn the Limits of Test-Time Compute for Self-EvaluationAnirudh Buvanesh, Khaled Rouissi, Sujin Yun, Johan Obando-Ceron +5
Self-verification and aggregation have emerged as popular approaches for bridging the gap between pass@1 and pass@N. However, increasing the number of candidates does not consistently narrow the gap between final-answer accuracy and pass@N. To understand this limitation, we use ranking error to measure how reliably models distinguish their correct generations from incorrect ones. We show that ranking error admits a bias-variance decomposition, with both components shaped by algorithmic design choices and inference compute. Across our experiments, additional compute reduces variance, while bias
benchmark - arxiv:2611.00344 · cs.ROSPARK: Self-Correcting Embodied Memory via Physical Re-ObservationDmytro Kurdydyk, Aadi Palnitkar, Sebastian Scherer, Seungchan Kim
Retrieval-augmented generation (RAG) systems have seen growing adoption in embodied settings as language-queryable memories of the environment. However, existing approaches often assume a reliable single-episode knowledge base (KB) or accumulate observations across episodes without correcting prior errors. Such errors can persist through incorrect semantic labels, inaccurate object positions, and inconsistent spatial relationships. To address this, we present SPARK, a corrective multi-episode framework for embodied RAG. After the first episode, a semantic critic flags regions of the KB, a targ
embodiedmemoryretrieval-augmented - arxiv:2611.00342 · cs.LGDEPLOY-RL: Active Boundary Discovery and Conservative Certification for Deployable Reinforcement Learning in Safety-Critical Continuous ProcessesYeojin Jang, Minu Baek, Gihun Gil, Minsung Jung +7
Reinforcement learning (RL) policies often outperform classical controllers in simulation, yet rarely reach production in safety-critical processes. The barrier is that there is no principled way to answer "Is this policy safe to deploy?" with statistical guarantees. We introduce DEPLOY-RL, a post-training certification framework built on one key insight: deployment certification requires discovering where failures occur (boundary discovery), not measuring how much everywhere (uniform reconstruction). Our contributions: (1) a contract-coupled acquisition function that concentrates sampling on
post-trainingbenchmarkevaluation protocol - arxiv:2611.00339 · cs.AIEfficient NPN Classification for Large Functions using Graph Neural NetworkChengdi Cao, Hanyu Wang, Wan-Hsuan Lin, Jason Cong
Two Boolean functions are considered NPN equivalent if one can be transformed into the other by negating inputs, permuting inputs, or negating the output. Efficient NPN classification is valuable in fields such as logic synthesis, quantum computing, and information theory. However, exact NPN classification demands exhaustive enumeration of all possible input permutations and negations. Its complexity grows exponentially, making it impractical for larger functions with more inputs. In this paper, we introduce the first learning-based formulation for NPN classification by recasting it as NPN act
benchmark - arxiv:2611.00333 · cs.LGC3: Speech-Guided KV Retrieval for Long-Context SpeechLLM InferenceDancheng Liu, Markus Müller, Rupak Vignesh Swaminathan, Jing Liu +2
Speech language models encode audio at a much higher rate than text, causing the KV cache to dominate the memory, bandwidth, and decoding cost of long-form generation. KV retrieval can reduce per-step access by attending to only a query-dependent subset of the cache, but existing methods are designed primarily for text and do not account for the structure of audio attention. We first show that retrieval errors may compound autoregressively into generation drift. We then analyze full audio attention during long-form generation and identify three properties relevant to retrieval: unequal attenti
long-context - arxiv:2611.00325 · cs.CLOne Waveform, Two Stories: Benchmarking Speech-Acoustic Conflict in Audio-Language ModelsWenbin Wang, Yuheng Li, Koichiro Yoshino
Audio-language models (ALMs) are evaluated on their understanding of both speech and environmental sounds. What remains underexplored is how they respond when spoken claims contradict acoustic evidence within the same recording. We introduce SAC-Bench, a benchmark of intra-audio speech--acoustic conflict comprising 616,031 unique clips from 188 sound-event classes across seven semantic, temporal, and spatial tasks. It pairs environmental recordings with compatible or contradictory synthesized spoken claims and includes audio-only controls where applicable. Across five open-weight ALMs evaluate
benchmark - arxiv:2611.00319 · cs.CLGAVAGAI: Referential Alignment for Infant-Scale Vision-Language Models via Balanced Optimal TransportRajdeep Pal, Jotiraditya Banerjee, Saptarshi Pani
A child learns which words name which things from roughly 100 million words of input. Vision-language models trained at a comparable scale bind words to referents far less reliably, and we argue that part of the gap is structural: the usual next-token captioning objective has no variable standing for "this word refers to that region," so training applies no direct pressure to ground words. We add such a variable by treating each scene as a small assignment problem between the content words heard and the regions seen, solved as entropic optimal transport under two constraints. First, a word may
benchmark - arxiv:2611.00314 · cs.AIDecentralized Multi-Agent Coordination and Communication in Shared CodebasesRishabh Mediratta, Lingyun Xiao, Peter Stone, Joydeep Biswas +1
Coding agents working in shared repositories may receive separate, independently-motivated requests from human developers, creating a setting that is multi-agent by necessity. When this scenario happens, their individual implementations may work separately but conflict when combined. To study this setting, we introduce ParallACT, a benchmark of orchestrator-free multi-agent coding tasks. ParallACT contains both complementary and contradictory tasks. In complementary tasks, each agent's task is separately useful and implementations must share an interface to be productively combined. In contrad
agentmulti-agentbenchmark - arxiv:2611.00312 · cs.LGRaMMOBO: Random Max-Margin Multi-Objective Bayesian OptimizationThomas Wedenig, Paul Plantosar, Robert Peharz
Multi-objective Bayesian optimization (MOBO) maximizes $M$ expensive black-box objectives with the goal of approximating the Pareto set, whose objective vectors cannot be improved in one component without sacrificing another. A common quality measure is hypervolume, the volume in objective space dominated by these vectors relative to a reference point. Expected hypervolume improvement (EHVI) directly targets this metric and is one-step Bayes-optimal, but its standard computation scales exponentially with $M$, making it prohibitively expensive for many-objective problems. We introduce random ma
benchmark - arxiv:2611.00299 · physics.app-phThe Agentic AI Kerr Lab: LLM Agents for Kerr Microscopy Experiments with Magnetic SkyrmionsKilian Leutner, Yuean Zhou, Aditya Kumar, Thomas B. Winkler +3
Magnetic skyrmions are topological spin textures that can be observed directly by magnetic microscopy and driven by fields and currents, which makes them a model system for fundamental physics and candidates for spintronic devices. Magneto-optical Kerr microscopy is a method to image them in real time; however, most experiments use an inefficient cycle of consecutive imaging, analysis and excitation in which every step requires the judgement of the experimentalist operating the setup. Here we present the Agentic AI Kerr Lab (AgentKerr), in which an agent based on a large language model operate
agentllm agentagentic - arxiv:2611.00297 · cs.ROTeach, Abstract, Apply: A Framework for Dialogue-Driven Task Learning with Collaborative RobotsChalamalasetti Kranti, Sherzod Hakimov, David Schlangen
Collaborative robots (cobots) are designed to be safer and more interactive than traditional industrial robots, share the same workspace as humans, and assist them in a range of shop-floor tasks such as palletization, joint assembly, and part inspection. Many of these tasks involve repetitive workflows that may evolve with changes in production environments and task specifications. These requirements motivate the need for intuitive interfaces that allow workers to modify workflows on the shop floor and generalise these routines for future use. This work introduces the Teach-Abstract-Apply (TAA
self-playhuman-in-the-loop - arxiv:2611.00292 · cs.AITADA: Technology Aware DNN Accelerator Performance and Energy EstimatorTanvi Sharma, Surya Selvam, Anand Raghunathan, Kaushik Roy
Deep neural network (DNN) accelerators for edge inference must meet tight energy, latency, and area (EDP) constraints, and their optimal hardware architecture and dataflow configuration is typically identified through design space exploration (DSE). However, on-chip memory technology including cell type, bank organization, and process node, is rarely treated as a first-class citizen in DSE. For example, increasing memory capacity can reduce memory accesses through greater data reuse, but at the cost of higher area and access energy, thereby limiting the compute units budget. Consequently, memo
memory - arxiv:2611.00291 · cs.AIWhen AI Agents Act on Our Behalf: A Taxonomy of Human-Agent Interaction FailureCanwen Wang, Qing Xiao, Haiyi Zhu, Hong Shen
AI agents write and execute code, manipulate user interfaces, and transact on their users' behalf. We ask what forms failure takes in human-agent interaction, and how those forms change as agents act with greater autonomy. From 695 documented incidents in two public repositories, we developed 12 failure modes bottom-up within three interaction stages adapted from Amershi et al.'s guidelines, coding the autonomy at which each arose. What a person can do when an agent fails depends on how much the agent completes between points of user involvement--a property fixed at deployment. Therefore, the
agentai agentagent benchmarkbenchmark - arxiv:2611.00290 · cs.AIQuantifying Local and Global Fidelity of Nonlinear EmbeddingsHenry Han
Dimension reduction methods are chosen by a common claim: t-SNE and UMAP preserve local structure, and PCA preserves global structure. Checking it on given data takes local and global scores on one scale and a test that compares two maps. Both scores come from one curve. The locality profile is the chance-corrected share of each point's k nearest neighbors that a map keeps, over all neighborhood sizes. Its value at k = 10 is the local fidelity L, and its mean over 30 to 50 percent of the data is the global fidelity G. We prove that at the midpoint the profile is the agreement on which half of
memoryagent memoryagentai agent - arxiv:2611.00288 · cs.CVSeeing Less in a Sparse World: Cross-View Geolocalization from Limited Field-of-View ObservationsFahimul Aleem, Shruti Vyas
Cross-view geolocalization aims to determine the geographic location of a ground-level image by retrieving its corresponding satellite image, offering a promising alternative for localization in Global Navigation Satellite System (GNSS)-denied or degraded environments such as urban canyons, dense forests, and disaster scenarios. Despite recent progress, existing approaches remain limited by the lack of globally diverse training datasets and benchmarks that are often geographically restricted. Furthermore, most benchmarks rely on panoramic ground views, which do not reflect the limited field-of
benchmark - arxiv:2611.00287 · cs.AIVerusWeave: A Benchmark for Proof Repair in Concurrent Verus ProgramsPedro Ribeiro, Martim Afonso, André Vasconcelos, João F. Ferreira
Concurrent software underpins critical infrastructure, including operating systems and databases, yet thread interactions can have bugs that bypass conventional testing. Verus extends Rust with formal verification capabilities, enabling developers to specify, implement and prove the correctness of concurrent code in the same language. Although coding agents have been explored for automated proof construction, existing Verus benchmarks offer limited coverage of concurrency-specific examples that connect atomic operations to abstract protocols. To address this gap, we introduce VerusWeave, the f
benchmark - arxiv:2611.00286 · cs.LGQuantization of Large Models with Weight Outliers and SparsityLeonidas Tsigkounakis, Abraham Jaeger Mountain, Yilang Zhang, Georgios B. Giannakis
Post-training quantization (PTQ) reduces the memory burden of large model deployment by casting the high precision trained weights to fewer bits. While this promotes memory-efficient inference, it also incurs quantization error that impairs model performance. To compensate for quantization error, the present contribution decomposes the original trained weights into a quantized matrix, a low-rank component, and a sparse matrix that accounts for outlier rows in the weights. Due to analytical intractability and the discrete structure of the decomposition, a simulated annealing search is employed
memorypost-trainingbenchmark - arxiv:2611.00284 · cs.LGOCELOT: Planetary-Scale Species Distribution Modeling with Satellite Foundation EmbeddingsBart van Merriënboer, Lauren Harrell, Keith Anderson, Burcu Karagol Ayan +3
Large-scale repositories of species observations now contain billions of occurrence records from around the world, but these sightings cluster heavily in human-accessible areas. Existing models use only a small fraction of these records and are not evaluated on extrapolating to remote, unvisited regions--where species distribution models can be used as geographic priors to help classify data from camera traps and acoustic recorders. We scale relative occurrence species distribution models by training a multimodal implicit neural representation network called OCELOT that fuses spatial coordinat
benchmark - arxiv:2611.00283 · cs.RODoes Validation Loss Predict Real-Robot Success? Evidence from Four Imitation-Learning Policies on a Low-Cost ArmSimon Cowling
Held-out validation loss is the standard offline signal for choosing checkpoints and comparing models, but its relation to closed-loop success on a real robot is rarely measured. We measure it for ACT, Diffusion Policy, SmolVLA and $π_{0.5}$ (action expert only), trained on nested subsets of 12 to 200 demonstrations of a pick-and-place task on an SO-101 arm. We score 82 checkpoints on 25 held-out demonstrations, with each policy's training loss and with the error of the actions it outputs, and compare the scores with 930 bench rollouts. We also score all 700 intermediate checkpoints of 70 trai
diffusion policy - arxiv:2611.00265 · cs.AIA generative foundation model for computational brain imagingYuting He, Xu Wen, Hongxing Lan, Chengliang Liu +9
Computational brain imaging is constrained by a mismatch between increasingly diverse tasks and scarce, heterogeneous clinical data. Although existing brain foundation models transfer pretrained representations, their knowledge remains entangled within model-specific features and fixed architectures. This paper proposes BrainGFM, a generative foundation model that extends knowledge transfer beyond representation reuse through a reusable and reconfigurable brain-imaging prior. By pretraining on a corpus of about 1.3 million MRI and CT slices, BrainGFM supported task-aligned data synthesis for l
post-training - arxiv:2611.00247 · cs.CVTaurus: LLM-in-the-Loop CT Annotation Matches Coding Agents at Half the CostBing Hu, Rishabh Sharma, Jatin Mehta
Dense labels of the aorta in CT are slow to draw, and the public aortic CT datasets that could stand in for them each label different structures under their own conventions. Large language model (LLM) agents can now write code and read images. We asked whether such a model labels these scans better as a free-form agent that writes its own segmentation code, or inside a harness, where an automated pass labels the scan and the model reviews the result. We built a harness, Taurus, and compared it with free-form agents by blind label recovery: we withheld the human labels of six public datasets (9
agent - arxiv:2611.00244 · cs.CLMEMENTO: Learning Context Management for Long Context Processing AgentsYunfan Zhang, Taehoon Hwang, Kathleen McKeown, Smaranda Muresan
LLM agents that operate over long inputs must perform context management: deciding what to retain in a fixed-size context window. A common approach is compaction, where the model replaces its history with a self-generated note. Yet off-the-shelf agentic models, including GPT-5.5 and GLM-5.3, often write ineffective notes, and reinforcement learning (RL) alone rarely discovers useful note-taking behavior. We present MEMENTO, a long-context agent built on Qwen3-8B and trained to manage its context through compaction. We first bootstrap note-taking by fine-tuning on synthetic trajectories annotat
long-contextlong contextagentllm agentagenticbenchmark - arxiv:2611.00242 · cs.AIRE-SciVQA: Improving Scientific VQA With Context-Relevant Retrieval and Lightweight MLLM AdaptationSheikh Tahmim Mehdi, Muminul Islam Jami
A deep understanding of scientific literature often relies on interpreting figures that contain key information beyond the main text. Leveraging Large Language Models (LLMs) to extract insights from these figures can significantly accelerate knowledge discovery. Existing works on Scientific Visual Question Answering rely on resource-intensive training pipelines that are often inaccessible and require large computational resources. Instead of designing complex architectures, we propose RE- SciVQA, which focuses on improving the quality of retrieved context and fine-tuning small LLMs with parame
ragbenchmark - arxiv:2611.00243 · cs.AIOmniPiano: Diverse Dexterous Piano-Playing Challenges for Standard, Robust, Safe, and Multi-Agent RLFan Xu, Fan Yang, Yifan Xu, Ruiyuan Zhang +6
Dexterous robotic manipulation remains a major challenge in robotics, exemplified by piano playing, which requires coordinated control of fingers and provides a demanding testbed for reinforcement learning (RL). Although previous work established a benchmark for robotic piano playing, it is restricted to two hands and does not support unified evaluation. To address this gap, we introduce OmniPiano, a benchmark that provides one to up to five Shadow Hands and supports standard, robust, safe, and multi-agent RL within a shared piano-playing task family. OmniPiano is developed for scalable and di
manipulationdexterousmulti-agentbenchmark - arxiv:2611.00236 · cs.LGProLand: A New Infeasible Manifold Optimization Method for LoRA Fine-TuningChenyu Zhou, Yuxuan Zeng, Qi Deng, Dongdong Ge +1
Low-rank adaptation (LoRA) enables parameter-efficient fine-tuning but can suffer from redundant basis directions. Constraining the basis factor to the Stiefel manifold addresses this redundancy through orthonormality. Landing methods allow temporarily infeasible iterates to explore a neighborhood of the manifold while progressively restoring orthogonality without per-step retractions. To improve optimization efficiency while retaining this flexibility, we introduce ProLand (Product-space Landing), a retraction-free optimizer that combines coordinated factor updates with Muon-style spectral di
benchmark - arxiv:2611.00234 · cs.LGLocalized anomaly detection with a calorimeter normalizing flowRaunav Mendiratta, Jason Weitz, Javier Duarte
Normalizing flows trained as fast calorimeter simulators also give the likelihood of each shower, which can serve as an anomaly score. Because this likelihood sums over every voxel, a signal confined to a few voxels can be lost in the fluctuations of the rest of the shower. For example, in a public benchmark dataset in which a long-lived neutral particle decays to two photons, the CaloFlow likelihood used in previous work selects less than 5% of the signal at a background efficiency of $10^{-3}$ when the particle has the largest simulated mass and decays at or just behind the front face of the
benchmark - arxiv:2611.00193 · cs.ROAnchor 1: Precise and Stable Humanoid Control for Whole-Body ManipulationArchon Robotics, :, Caojun Wang, Checheng Yu +23
Humanoid controllers have acquired increasingly diverse motor skills, yet translating these capabilities into precise and stable whole-body manipulation remains challenging. Manipulation tasks are sensitive to execution deviations -- even small errors or unintended steps can cause failure. Introducing Anchor 1, a whole-body controller for precise and stable whole-body manipulation, together with a benchmark spanning diverse coordination demands. Anchor 1 adopts calibrated sparse motion references to preserve task-relevant human motion structure without online joint-space retargeting, leaving k
manipulationhumanoidwhole-body controlbenchmark - arxiv:2611.00192 · cs.CVVC-1K: A Diverse Video Test Set for Real-Time CommunicationYu-Chen Sun, Jie Dong, Sergey Sablin, Hassene Tmar +1
We introduce Video Conferencing 1000 (VC-1K), an open-source test set comprising approximately 1,000 video sequences specifically designed for evaluating video processing and encoding pipelines in real-time communication (RTC) scenarios. Unlike existing test sets that primarily rely on studio-produced or screen content, VC-1K consists of user-generated video call content captured on consumer mobile devices under a wide variety of conditions. The dataset predominantly features talking-head sequences with distinct faces, but also includes a small subset of outward-facing, rear-camera videos with
benchmark - arxiv:2611.00191 · cs.CVMemU-Bench: Benchmarking GUI Agents on Long-Horizon User InteractionsYizhou Liu, Zhengxi Lu, Fei Tang, Yiwen Qiu +11
Existing Graphical User Interface (GUI) benchmarks predominantly rely on precise and stateless user instructions. In everyday usage, however, instructions are often \emph{vague and history-dependent}: users omit the exact details of their past operations because they no longer remember them, leaving the agent to recover the missing context from long and noisy interaction histories. To bridge this evaluation gap, we introduce \textbf{MemU-Bench}, a mobile benchmark of \textbf{118 vague tasks} grounded in 29 long-horizon user trajectories, where agents must distill raw interaction streams into m
memoryagentbenchmark - arxiv:2611.00186 · cs.AINeurons Are All You NeedJoseph B. Alzagatiti, Alain C. Glanzman, Perry A. Demsko, Pablo I. Salanova +1
For decades, artificial neural networks have operated under the assumption that durable, learned capability resides exclusively in adaptive synaptic parameters, primarily due to the established neurobiology of synaptic connectivity. This design forces models to remain stateless during deployment, incurring well-known limitations such as catastrophic forgetting and expensive retraining. However, recent neurobiological research by Glanzman and colleagues, and others, shows that memories can be transferred between individual organisms in a synaptically independent manner. Inspired by this researc
memorypersistent memory - arxiv:2611.00180 · cs.ROLearning to Live: A Step Toward Lifelong Embodied IntelligenceLirong Che, Peidong Wang, Yimeng Su, Jiong Wang +10
Embodied agents are usually organized around individual tasks, yet the physical world keeps changing between tasks and during execution. We introduce LIFE (Lifelong Intelligence from Experience), a framework in which a robot observes continuously, remembers selectively, and turns experience into new capabilities. RoboBeats, a family of lightweight streaming gates, tracks conditions set by the agent and discovers new events, deciding when an observation warrants a response or a memory write. Selected observations enter Plastic Memory, a flat store of images and structured records that keeps the
embodiedmemoryagentembodied agent - arxiv:2611.00177 · cs.CVAuroraWorld: Structured 4D World Modeling for Exo-to-Ego Video Generation in Robot ManipulationHao Xue, Peng-Fei Zhang, Ying Cheng, Lanyun Zhu +2
Egocentric observations provide a robot-centric view of manipulation processes, capturing contact interactions and object-state changes that are important for visuomotor policy learning but difficult to observe from external cameras. However, collecting such data at scale remains expensive due to additional hardware requirements and platform-specific deployment costs. Synthesizing egocentric observations from large-scale exocentric manipulation videos provides a scalable alternative. However, existing methods struggle to synthesize physically consistent and policy-relevant egocentric videos un
manipulationworld model - arxiv:2611.00171 · cs.AIReimagining Text-to-SQL: A Survey of Large Language Model Architectures and StrategiesShikhar Dave, Bhaskarjit Sarmah, Stefano Pasquali
Text-to-SQL systems translate natural language questions into executable SQL queries, democratizing access to relational data. Early approaches relied on handcrafted rules and template matching, while subsequent neural methods introduced encoder-decoder architectures, schema linking, and constrained decoding to handle increasingly complex queries. The emergence of large language models (LLMs), exemplified by systems such as ReFoRCE, Agentar-Scale-SQL, Snowflake's Arctic-Text2SQL-R1, and the modular XiYan-SQL framework, has dramatically advanced the field, pushing execution accuracy to new stat
retrieval-augmentedagenticbenchmark - arxiv:2611.00169 · cs.AIPatchScout: Thematic Web Data Collection via Information ForagingMichael West, Eduard Dragut
Researchers often need web corpora that are not answers to a single query, but reusable collections of documents spanning many decentralized sources. We refer to this problem as thematic web data collection: assembling semantically relevant documents that share a common theme but are scattered across heterogeneous and structurally disconnected regions of the web. Existing approaches, including traversal-based methods and LLM-driven web agents, are limited by local exploration or short-horizon stopping, preventing comprehensive collection. We formulate this task as a long-horizon information fo
agentmulti-agentagent framework - arxiv:2611.00165 · cs.LGLearning to bid and bundle: a posterior-sampling agent for online retailHadi Mohammadi
An online retailer launching a perishable product bids for sponsored-search clicks, discounts a bundle with an always-stocked staple and displays its remaining stock, which raises willingness to pay. It knows neither how clicks respond to its bid nor how strong and dispersed its market is, and learns both over repeated seasons from data its own decisions censor: a zero bid reveals nothing about paid clicks, and choices follow only clicks. We design a posterior-sampling agent for it. Each period it draws a market from an exact posterior and posts the bid and discount of that market's exactly so
agent - arxiv:2611.00159 · cs.LGA Generalized Sparse Fuzzy Clustering Framework with Automated Cluster Cardinality Estimation AbilityGeorge Stamatelis, Konstantinos Koutroumbas
Fuzzy clustering provides a robust framework for handling ambiguous and overlapping data boundaries, yet its traditional formulation generates entirely dense membership assignments that limit noise resilience and prevent estimating the true number of clusters, when it is unknown. To address these limitations, we introduce in this paper the Generalized Sparse Fuzzy Clustering (GSFC) framework, which incorporates a continuous sparsity penalty into the fuzzy objective function. Because closed-form solutions are precluded by the simultaneous presence of fuzzifier and sparsity exponents, we propose
benchmark - arxiv:2611.00158 · cs.ROScaling Trusted Autonomy: Towards Generally Capable Verifiable RobotsBenedict Quartey
Robots powered by foundation models are increasingly able to interpret open-ended language, recognize unfamiliar objects, and operate across diverse environments. Yet broader capability does not establish that a robot has understood a request, that its actions will produce the intended effects, or that its internal representation still corresponds to reality. Classical symbolic planning provides explicit structure for composition and verification, but traditionally depends on hand-engineered skills and representations. Learned policies can broaden the robot's repertoire, but provide weak guara
manipulation - arxiv:2611.00157 · cs.LGFlowFPR: Flow Matching Prior for Oversampled Fourier Phase RetrievalLi Jiatai, Cao Wenfei, Chen Yang
Fourier phase retrieval reconstructs images from magnitude measurements without phase information, but noise and channel ambiguities make natural color-image recovery challenging. Although diffusion-prior methods provide strong image constraints, their recovery can deteriorate under stronger measurement noise, and repeated generative updates incur substantial computational cost. This paper proposes FlowFPR, which integrates a pretrained Flow Matching prior with relaxed averaged alternating reflections for oversampled Fourier phase retrieval. Each outer iteration obtains a clean image estimate
evaluation protocol - arxiv:2611.00155 · cs.ROFrom Concepts to Actual Causes: Toward Closing the Concept-Malfunction Gap in Robot Foundation ModelsGihyuk Ko
Post-incident analysis of a robot foundation model poses a singular causal question: which internal components caused this failure, and to what degree? Mechanistic interpretability (MI), however, establishes general causal claims by decoding the concepts a policy has learned, yet knowing which concepts are encoded is insufficient for identifying the actual cause of a specific incident. In this position paper, we analyze this concept-malfunction gap by posing fault isolation in the framework of actual causality. Specifically, we formalize a fault-isolation query as a tuple of intervened interna
robot foundation model
02 US SEMI · SEC 8-K FILINGS
8 itemsscanned: NVDA / AVGO / MRVL / COHR / LITE / AMD / TSM / SMCI / ANET / CRDO / POWL / VECO
- $ANET · 8-K · filed 2026-11-03Arista Networks IncItems: 2.02,9.018-K
- $SMCI · 8-K · filed 2026-11-03Super Micro Computer IncItems: 2.02,8.01,9.018-K
- $AMD · 8-K · filed 2026-11-03Advanced Micro Devices IncItems: 2.02,7.01,9.018-K
- $LITE · 8-K · filed 2026-11-02Lumentum Holdings IncItems: 2.02,9.018-K
- $LITE · 8-K · filed 2026-11-02Lumentum Holdings IncItems: 1.01,5.02,7.01,9.018-K
- $MRVL · 8-K · filed 2026-11-02Marvell Technology IncItems: 1.01,9.018-K
- $MRVL · 8-K · filed 2026-11-02Marvell Technology IncItems: 2.02,7.01,9.018-K
- $VECO · 8-K · filed 2026-11-02Veeco Instruments IncItems: 7.01,8.01,9.01FORM 8-K
03 HUMANOID · COMPANY NEWS
62 itemsscanned: figure-ai / 1x / boston-dynamics / unitree / apptronik / sanctuary-ai / neura-robotics / agility-robotics / physical-intelligence / agibot
Figure AI (10)
- Figure AISeptember 30, 2026F.02 Decommission
- Figure AISeptember 17, 2026Helix 2.5: Zero-Shot 30-Home Generalization
- Figure AIOctober 09, 2025Introducing Figure 03
- Figure AISeptember 03, 2026Figure and Nscale Sign Strategic Partnership For Up to 100,000 GPUs on the NVIDIA Vera Rubin Platform
- Figure AIAugust 25, 2026Introducing Index: Building The World’s Largest and Most Diverse Physical Dataset
Boston Dynamics (10)
Unitree 宇树 (10)
- Unitree 宇树Components
- Unitree 宇树Kung Fu Meets Spring, Unitree SFG Robots Present "Cyber Real Kung Fu" in the Year of the Horse2026-05-31Media Coverage
- Unitree 宇树Welcoming Myanmar President Min Aung Hlaing to Unitree2026-08-05Media Coverage
- Unitree 宇树Unitree founder Wang Xingxing graces the cover of Time magazine2026-08-05Media Coverage
- Unitree 宇树Unitree Announces H2 Plus, an NVIDIA Isaac GR00T Reference Humanoid Robot for Academic Research2026-06-01Media Coverage
Apptronik (1)
Sanctuary AI (6)
- Sanctuary AIBlog PostTIGER: Task-Space Imitation Guidance for Efficient Reinforcement LearningDemonstrations as a compass, not a scriptRead More
- Sanctuary AIPress ReleaseProduct UpdatesSanctuary AI Expands Physical AI Strategy to Industrial Robotics, Demonstrating Production-Ready AI PerformanceRead More
- Sanctuary AICorporate NewsDaniel Friedmann Appointed CEO of Sanctuary AIRead More
- Sanctuary AIPress ReleaseZeon Invests in Sanctuary AI and Partners to Advance Specialized Materials for Dexterous RoboticsRead More
- Sanctuary AIThought LeadershipWeb Summit Reflections: Canada’s Physical AI Moment Can’t WaitRead More
Agility Robotics (10)
- Agility RoboticsAgility’s Humanoid Deployment ProcessPras VelagapudiAugust 04, 2026
- Agility RoboticsJonathan Hurst answers your questions about building Digit 5InsightOctober 23, 2026
- Agility RoboticsThe Realistic Pathway to HomeInsightMay 26, 2026
- Agility RoboticsAgility and AIInsightMarch 16, 2026
- Agility RoboticsAgility Gets a New BrandInsightMarch 5, 2026
Physical Intelligence (8)
- Physical IntelligenceFrontier Models for Every RobotOctober 13, 2026Our latest robotics models are now available through the Pi API beta. Upload data, fine-tune models, and deploy cloud-hosted inference on your own robots.
- Physical Intelligenceπ0.7: a Steerable Model with Emergent CapabilitiesApril 16, 2026A steerable robotic foundation model that exhibits a step-change in generalization.
- Physical IntelligenceThe Physical Intelligence LayerFebruary 24, 2026General-purpose physical intelligence models will enable a Cambrian explosion of robotics applications. See how our partners are already solving real-world problems.
- Physical IntelligenceMoravec's Paradox and the Robot OlympicsDecember 22, 2025By fine-tuning our latest model, we were able to solve a series of very difficult manipulation challenge tasks.
- Physical Intelligenceπ*0.6: a VLA that Learns from ExperienceNovember 17, 2025A method for training our generalist policies with RL to improve success rate and throughput on real-world tasks.
智元 AgiBot (7)
- 智元 AgiBotAGIBOT Unveils GO-3 Preview, Takes Top Spot on Three Embodied AI Benchmarks2026-10-29
- 智元 AgiBotHow Much Work Do Robots Really Deliver? The Thinking Behind Worken2026-10-29
- 智元 AgiBotAGIBOT Introduces Worken to Measure Productive Robot Work at AGIBOT Innovation Day 20262026-10-28
- 智元 AgiBotAGIBOT Launches Genie Studio 2.0 to Bring Robots into Real Workplaces2026-10-26
- 智元 AgiBotAGIBOT Open-Sources AGIBOT WORLD Release 4: "Physical Reasoning" Dataset Enabling Robots to Reason Before Acting2026-10-23