Weekly Digest — 2025-W29

Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflows - all through natural language commands.

Python tool for converting files and office documents to Markdown.

uBlock Origin - An efficient blocker for Chromium and Firefox. Fast and lean.

Qlib is an AI-oriented Quant investment platform that aims to use AI tech to empower Quant Research, from exploring ideas to implementing productions. Qlib supports diverse ML modeling paradigms, including supervised learning, market dynamics modeling, and RL, and is now equipped with https://github.com/microsoft/RD-Agent to automate R&D process.

Next.js Commerce

AI's query engine - Platform for building AI that can answer questions over large scale federated data. - The only MCP Server you'll ever need

FULL v0, Cursor, Manus, Same.dev, Lovable, Devin, Replit Agent, Windsurf Agent, VSCode Agent, Dia Browser, Trae AI & Cluely (And other Open Sourced) System Prompts, Tools & AI Models.

Build Real-Time Knowledge Graphs for AI Agents

CUDA Templates for Linear Algebra Subroutines

Free and Open Source Enterprise Resource Planning (ERP)

Use Wireless Android Auto with a car that supports only wired Android Auto using a Raspberry Pi.

Chat with your documents on your local device using GPT models. No data leaves your device and 100% private.

Turn anything into ready-to-post videos with one click

Local-first AI builder for mobile and web apps

Make, share, swipe

Find tech internships in your favorite startups!

Cloud software for rental businesses

Ultra-realistic text-to-speech

Connect direct to Salesforce & HubSpot with native formulas

Let your AI code — we’ll make it work.

AI accounting for founders & accountants

Imagine Cursor, but for PowerPoint presentations

Monitor and optimize brand presence across AI platforms

The open-source framework for real-time AI voice

Retrieval-Augmented Generation (RAG) lifts the factuality of Large Language Models (LLMs) by injecting external knowledge, yet it falls short on problems that demand multi-step inference; conversely, purely reasoning-oriented approaches often hallucinate or mis-ground facts. This survey synthesizes both strands under a unified reasoning-retrieval perspective. We first map how advanced reasoning optimizes each stage of RAG (Reasoning-Enhanced RAG). Then, we show how retrieved knowledge of different type supply missing premises and expand context for complex inference (RAG-Enhanced Reasoning). Finally, we spotlight emerging Synergized RAG-Reasoning frameworks, where (agentic) LLMs iteratively interleave search and reasoning to achieve state-of-the-art performance across knowledge-intensive benchmarks. We categorize methods, datasets, and open challenges, and outline research avenues toward deeper RAG-Reasoning systems that are more effective, multimodally-adaptive, trustworthy, and human-centric. The collection is available at https://github.com/DavidZWZ/Awesome-RAG-Reasoning.

3D modeling is moving from virtual to physical. Existing 3D generation primarily emphasizes geometries and textures while neglecting physical-grounded modeling. Consequently, despite the rapid development of 3D generative models, the synthesized 3D assets often overlook rich and important physical properties, hampering their real-world application in physical domains like simulation and embodied AI. As an initial attempt to address this challenge, we propose PhysX, an end-to-end paradigm for physical-grounded 3D asset generation. 1) To bridge the critical gap in physics-annotated 3D datasets, we present PhysXNet - the first physics-grounded 3D dataset systematically annotated across five foundational dimensions: absolute scale, material, affordance, kinematics, and function description. In particular, we devise a scalable human-in-the-loop annotation pipeline based on vision-language models, which enables efficient creation of physics-first assets from raw 3D assets.2) Furthermore, we propose PhysXGen, a feed-forward framework for physics-grounded image-to-3D asset generation, injecting physical knowledge into the pre-trained 3D structural space. Specifically, PhysXGen employs a dual-branch architecture to explicitly model the latent correlations between 3D structures and physical properties, thereby producing 3D assets with plausible physical predictions while preserving the native geometry quality. Extensive experiments validate the superior performance and promising generalization capability of our framework. All the code, data, and models will be released to facilitate future research in generative physical AI.

Humans are integral components of the transportation ecosystem, and understanding their behaviors is crucial to facilitating the development of safe driving systems. Although recent progress has explored various aspects of human behaviorx2014such as motion, trajectories, and intentionx2014a comprehensive benchmark for evaluating human behavior understanding in autonomous driving remains unavailable. In this work, we propose MMHU, a large-scale benchmark for human behavior analysis featuring rich annotations, such as human motion and trajectories, text description for human motions, human intention, and critical behavior labels relevant to driving safety. Our dataset encompasses 57k human motion clips and 1.73M frames gathered from diverse sources, including established driving datasets such as Waymo, in-the-wild videos from YouTube, and self-collected data. A human-in-the-loop annotation pipeline is developed to generate rich behavior captions. We provide a thorough dataset analysis and benchmark multiple tasksx2014ranging from motion prediction to motion generation and human behavior question answeringx2014thereby offering a broad evaluation suite. Project page : https://MMHU-Benchmark.github.io.

Large Language Model (LLM) agents have shown great potential for solving real-world problems and promise to be a solution for tasks automation in industry. However, more benchmarks are needed to systematically evaluate automation agents from an industrial perspective, for example, in Civil Engineering. Therefore, we propose DrafterBench for the comprehensive evaluation of LLM agents in the context of technical drawing revision, a representation task in civil engineering. DrafterBench contains twelve types of tasks summarized from real-world drawing files, with 46 customized functions/tools and 1920 tasks in total. DrafterBench is an open-source benchmark to rigorously test AI agents' proficiency in interpreting intricate and long-context instructions, leveraging prior knowledge, and adapting to dynamic instruction quality via implicit policy awareness. The toolkit comprehensively assesses distinct capabilities in structured data comprehension, function execution, instruction following, and critical reasoning. DrafterBench offers detailed analysis of task accuracy and error statistics, aiming to provide deeper insight into agent capabilities and identify improvement targets for integrating LLMs in engineering applications. Our benchmark is available at https://github.com/Eason-Li-AIS/DrafterBench, with the test set hosted at https://huggingface.co/datasets/Eason666/DrafterBench.

Code performance optimization is paramount in real-world software engineering and critical for production-level systems. While Large Language Models (LLMs) have demonstrated impressive capabilities in code generation and bug fixing, their proficiency in enhancing code performance at the repository level remains largely unexplored. To address this gap, we introduce SWE-Perf, the first benchmark specifically designed to systematically evaluate LLMs on code performance optimization tasks within authentic repository contexts. SWE-Perf comprises 140 carefully curated instances, each derived from performance-improving pull requests from popular GitHub repositories. Each benchmark instance includes the relevant codebase, target functions, performance-related tests, expert-authored patches, and executable environments. Through a comprehensive evaluation of representative methods that span file-level and repo-level approaches (e.g., Agentless and OpenHands), we reveal a substantial capability gap between existing LLMs and expert-level optimization performance, highlighting critical research opportunities in this emerging field.

Enabling virtual humans to dynamically and realistically respond to diverse auditory stimuli remains a key challenge in character animation, demanding the integration of perceptual modeling and motion synthesis. Despite its significance, this task remains largely unexplored. Most previous works have primarily focused on mapping modalities like speech, audio, and music to generate human motion. As of yet, these models typically overlook the impact of spatial features encoded in spatial audio signals on human motion. To bridge this gap and enable high-quality modeling of human movements in response to spatial audio, we introduce the first comprehensive Spatial Audio-Driven Human Motion (SAM) dataset, which contains diverse and high-quality spatial audio and motion data. For benchmarking, we develop a simple yet effective diffusion-based generative framework for human MOtion generation driven by SPatial Audio, termed MOSPA, which faithfully captures the relationship between body motion and spatial audio through an effective fusion mechanism. Once trained, MOSPA could generate diverse realistic human motions conditioned on varying spatial audio inputs. We perform a thorough investigation of the proposed dataset and conduct extensive experiments for benchmarking, where our method achieves state-of-the-art performance on this task. Our model and dataset will be open-sourced upon acceptance. Please refer to our supplementary video for more details.

The large language model (LLM) community focuses almost exclusively on decoder-only language models, since they are easier to use for text generation. However, a large subset of the community still uses encoder-only models for tasks such as classification or retrieval. Previous work has attempted to compare these architectures, but is forced to make comparisons with models that have different numbers of parameters, training techniques, and datasets. We introduce the SOTA open-data Ettin suite of models: paired encoder-only and decoder-only models ranging from 17 million parameters to 1 billion, trained on up to 2 trillion tokens. Using the same recipe for both encoder-only and decoder-only models produces SOTA recipes in both categories for their respective sizes, beating ModernBERT as an encoder and Llama 3.2 and SmolLM2 as decoders. Like previous work, we find that encoder-only models excel at classification and retrieval tasks while decoders excel at generative tasks. However, we show that adapting a decoder model to encoder tasks (and vice versa) through continued training is subpar compared to using only the reverse objective (i.e. a 400M encoder outperforms a 1B decoder on MNLI, and vice versa for generative tasks). We open-source all artifacts of this study including training data, training order segmented by checkpoint, and 200+ checkpoints to allow future work to analyze or extend all aspects of training.

The performance of Large Language Models (LLMs) is fundamentally determined by the contextual information provided during inference. This survey introduces Context Engineering, a formal discipline that transcends simple prompt design to encompass the systematic optimization of information payloads for LLMs. We present a comprehensive taxonomy decomposing Context Engineering into its foundational components and the sophisticated implementations that integrate them into intelligent systems. We first examine the foundational components: context retrieval and generation, context processing and context management. We then explore how these components are architecturally integrated to create sophisticated system implementations: retrieval-augmented generation (RAG), memory systems and tool-integrated reasoning, and multi-agent systems. Through this systematic analysis of over 1300 research papers, our survey not only establishes a technical roadmap for the field but also reveals a critical research gap: a fundamental asymmetry exists between model capabilities. While current models, augmented by advanced context engineering, demonstrate remarkable proficiency in understanding complex contexts, they exhibit pronounced limitations in generating equally sophisticated, long-form outputs. Addressing this gap is a defining priority for future research. Ultimately, this survey provides a unified framework for both researchers and engineers advancing context-aware AI.

Recent advancements in vision-language models (VLMs) have improved performance by increasing the number of visual tokens, which are often significantly longer than text tokens. However, we observe that most real-world scenarios do not require such an extensive number of visual tokens. While the performance drops significantly in a small subset of OCR-related tasks, models still perform accurately in most other general VQA tasks with only 1/4 resolution. Therefore, we propose to dynamically process distinct samples with different resolutions, and present a new paradigm for visual token compression, namely, VisionThink. It starts with a downsampled image and smartly decides whether it is sufficient for problem solving. Otherwise, the model could output a special token to request the higher-resolution image. Compared to existing Efficient VLM methods that compress tokens using fixed pruning ratios or thresholds, VisionThink autonomously decides whether to compress tokens case by case. As a result, it demonstrates strong fine-grained visual understanding capability on OCR-related tasks, and meanwhile saves substantial visual tokens on simpler tasks. We adopt reinforcement learning and propose the LLM-as-Judge strategy to successfully apply RL to general VQA tasks. Moreover, we carefully design a reward function and penalty mechanism to achieve a stable and reasonable image resize call ratio. Extensive experiments demonstrate the superiority, efficiency, and effectiveness of our method. Our code is available at https://github.com/dvlab-research/VisionThink.

We introduce pi^3, a feed-forward neural network that offers a novel approach to visual geometry reconstruction, breaking the reliance on a conventional fixed reference view. Previous methods often anchor their reconstructions to a designated viewpoint, an inductive bias that can lead to instability and failures if the reference is suboptimal. In contrast, pi^3 employs a fully permutation-equivariant architecture to predict affine-invariant camera poses and scale-invariant local point maps without any reference frames. This design makes our model inherently robust to input ordering and highly scalable. These advantages enable our simple and bias-free approach to achieve state-of-the-art performance on a wide range of tasks, including camera pose estimation, monocular/video depth estimation, and dense point map reconstruction. Code and models are publicly available.

Length generalization, the ability to solve problems of longer sequences than those observed during training, poses a core challenge of Transformer-based large language models (LLM). Although existing studies have predominantly focused on data-driven approaches for arithmetic operations and symbolic manipulation tasks, these approaches tend to be task-specific with limited overall performance. To pursue a more general solution, this paper focuses on a broader case of reasoning problems that are computable, i.e., problems that algorithms can solve, thus can be solved by the Turing Machine. From this perspective, this paper proposes Turing MAchine Imitation Learning (TAIL) to improve the length generalization ability of LLMs. TAIL synthesizes chain-of-thoughts (CoT) data that imitate the execution process of a Turing Machine by computer programs, which linearly expands the reasoning steps into atomic states to alleviate shortcut learning and explicit memory fetch mechanism to reduce the difficulties of dynamic and long-range data access in elementary operations. To validate the reliability and universality of TAIL, we construct a challenging synthetic dataset covering 8 classes of algorithms and 18 tasks. Without bells and whistles, TAIL significantly improves the length generalization ability as well as the performance of Qwen2.5-7B on various tasks using only synthetic data, surpassing previous methods and DeepSeek-R1. The experimental results reveal that the key concepts in the Turing Machine, instead of the thinking styles, are indispensable for TAIL for length generalization, through which the model exhibits read-and-write behaviors consistent with the properties of the Turing Machine in their attention layers. This work provides a promising direction for future research in the learning of LLM reasoning from synthetic data.

Controllable captioning is essential for precise multimodal alignment and instruction following, yet existing models often lack fine-grained control and reliable evaluation protocols. To address this gap, we present the AnyCap Project, an integrated solution spanning model, dataset, and evaluation. We introduce AnyCapModel (ACM), a lightweight plug-and-play framework that enhances the controllability of existing foundation models for omni-modal captioning without retraining the base model. ACM reuses the original captions from base models while incorporating user instructions and modality features to generate improved captions. To remedy the data scarcity in controllable multimodal captioning, we build AnyCapDataset (ACD), covering three modalities, 28 user-instruction types, and 300\,k high-quality data entries. We further propose AnyCapEval, a new benchmark that provides more reliable evaluation metrics for controllable captioning by decoupling content accuracy and stylistic fidelity. ACM markedly improves caption quality across a diverse set of base models on AnyCapEval. Notably, ACM-8B raises GPT-4o\'s content scores by 45\% and style scores by 12\%, and it also achieves substantial gains on widely used benchmarks such as MIA-Bench and VidCapBench.

比特币币值周一突破了 12 万美元（最高 122,571.19 美元，之后小幅回落），创历史新高。美国众议院正在制定加密货币资产相关的国家监管框架，而自称加密货币总统的特朗普则督促议员制定有利于加密货币行业的政策。分析师表示多项利好因素推动了币值上涨。比特币今年迄今已上涨 29%，它的涨势也推动了其它加密货币币值的上涨。

旨在帮助用户管理磁盘分区的发行版 GParted Live 释出了 v1.7.0。新版本的一大变化是停止支持 32 位架构，原因是 GParted Live 是基于 Debian Sid，而 Debian 已从 Sid 软件包库中移除了 i386 kernel 包，因此它只提供 64 位 amd64 版本。GParted Live 1.7.0 的其他变化包括：GParted 1.7.0，基于 Debian Sid (截至 2025/Jul/12），Linux 内核版本更新到 6.12.37-1，实现了避免启动时块设备随机排序的机制——此举旨在防止用户在多个驱动器的系统中选择错误的磁盘。

由于无人机技术的快速创新，乌克兰战争前线日益陷入僵局。交战双方都有数百架无人机在 1200 公里长的前线上空盘旋。无人机能将各种物质，包括食物、水、弹药、移动电源甚至偶尔还有灭火器，运送到前线，士兵们无需面临对方无人机的袭击而去穿越战场中最危险的部分。在战争中，没有一项创新会比第一人称视角（FPV）无人机影响更大。FPV 无人机能捆绑炸药，直接飞向目标，变成低成本自杀式炸弹。虽然爆炸威力不如火箭弹，但精度更高，且制造成本低廉能大规模部署。俄罗斯也在乌克兰之后开始采用 FPV 无人机。FPV 无人机的大规模使用对减缓前线的移动速度起到了关键作用。FPV 无人机很难被击落，主要防御手段是无线电干扰。虽然大部分无人机创新源自乌克兰，但俄罗斯首先实现了 FPV 无人机的一项重要改进：为其引入了连接无人机和操作人员的光纤去对抗干扰。

英伟达发布安全通知，警告其高端 GPU 使用的 GDDR6 显存在未启用 ECC 的情况下易受 RowHammer 比特翻转攻击。RowHammer 攻击指的是通过反复访问内存芯片的特定区域导致比特翻转。比特翻转（Bitflips）是指储存在电子设备上的个别比特发生翻转的事件，比如从 0 变为 1 或反之亦然。 RowHammer 比特翻转攻击已经存在了很多年，英伟达建议如果 GPU 支持 ECC 就启用该功能。GDDR7 和 HBM3 内置了 ECC 因此能自动抵御 RowHammer 攻击。

在论文截至前一周，大四学生 Xiaobing 收到了学校的通知，称论文如果有三成以上内容被标记为 AI 生成将会遭到拒绝。Xiaobing 表示论文都是自己写的，只有几段用 ChatGPT 和 DeepSeek 润色下。出于安全起见，她花了 70 元在学校计划使用的 AI 测试平台测试了下，结果 AI 检测器声称论文一半内容是 AI 生成的。她倍感震惊。有无数学生面临类似的问题，这一情况迫使他们用 AI 去反制 AI 检测。知网、万方数据和维普等学术数据库既向学校出售 AI 检测工具，也向学生出售打败 AI 检测的工具，从两方收钱。部分学生花数百元去润色论文以通过 AI 检测，但结果好坏参半。润色者被发现使用的也是 AI。一位学生称，AI 辅助服务将半导体润色为“0.5 导体”。

美国电动汽车制造商 Lucid Motors 创造了单次充电行驶里程 1205 公里的吉尼斯世界记录。 Lucid Motors 的电动汽车 Air Grand Touring 完成了从瑞士 St. Moritz 到德国慕尼黑的旅程，途经高速公路、二级公路和高山公路，期间没有停车充电。Air Grand Touring 配备了一块 NMC 电池，总容量 117 kWh，可用 112 kWh，能在三秒内从 0 加速到 60 mph。它的起售价 112,650 美元，是市场上最奢华的汽车之一。

高档车型使用的车载芯片通常来自英伟达和高通，但在普通车型使用性价比更高的中国半导体厂商日益受到青睐。截至 2024 年 9月 底，成立于 2015 年的地平线机器人公司的 SoC 已经获得 27 家汽车厂商的 285 款车型采用。地平线的下一个目标是走向世界。其战略是与欧洲的大型供应商联手，将自己的 SoC 销往世界市场。它的合作伙伴包括了德国博世（Bosch）和德国大陆集团（Continental）。另一家公司芯驰科技的 SoC 被本田和日产汽车的中国合资公司在其车型中采用。德国大众（VW）集团在巴西和印度等地销售的新车型采用了吉利控股集团旗下的芯擎科技开发的座舱用 SoC“龙鹰一号”。地缘政治风险是中国车载芯片厂商全球化计划面临的巨大挑战，至少美国汽车厂商不太可能会使用中国的 SoC。

xAI 的新模型 Grok 4 被发现在回答敏感问题前会先检查马斯克（Elon Musk）的观点。在这之前 Grok 一度宣称自己是机械希特勒（MechaHitler）。AI 研究员 Simon Willison 问 Grok 在巴以冲突中它站在哪一方？模拟推理过程的“思维轨迹（thinking trace）”显示它搜索了 X 平台上马斯克的帖子——from:elonmusk (Israel OR Palestine OR Gaza OR Hamas)，然后回答以色列。Grok 称鉴于马斯克的影响力其立场可以作为参考。Grok 并不总是会去搜索马斯克的帖子，因此 Willison 猜测 Grok 的这种行为是基于一系列推理：它知道 Grok 4 是 xAI 构建的，马斯克是 xAI 的所有者, 因此在寻求建议时它通常会更多考虑马斯克的想法。

2012 年独立安全研究员 Neil Smith 向美国政府报告了列车通信标准的漏洞，他可能没想到，直到 2025 年美国政府才披露了该漏洞。该漏洞允许攻击者使用软件定义无线电远程对火车进行刹车。US Cybersecurity and Infrastructure Security Agency (CISA)公布了名为 CVE-2025-1727 的漏洞，该漏洞与车尾到车头的链路协议弱验证有关，货车尾部的 Flashing Rear-End Device（FRED）设备向机头传输数据的系统使用了 BCH 校验和创建数据包，容易被软件定义无线电嗅探然后伪造数据包，向 FRED 施加制动，有可能引发脱轨事故。目前该问题尚未修复，美国铁路协会表示计划部署更安全的替代方案，但可能要到 2027 年才会推出。

英伟达宣布恢复向中国出口 H20 芯片，并表示将会推出一款特供中国市场的新型号芯片。H20 此前是英伟达能向中国出口的最先进 AI 芯片，美国政府于 4 月 9 日通知英伟达，H20 芯片需要获得许可证才能出口到中国。英伟达现在表示它已经申请了许可，已经获得政府承诺，预计将会很快获得批准，准备开始向中国公司交付芯片。英伟达同时表示，正在为中国市场开发一款新的特供 AI 芯片，完全符合美国的出口管制规定。根据早些时候的报道，新芯片将是基于 Blackwell 架构的 RTX Pro 6000D，使用 GDDR7 而不是 HBM 等更先进的高带宽显存，预计售价在 6,500-8,000 美元之间，低于 H20 的 10,000-12,000 美元。

成本上升和油价下跌迫使沙特缩减其雄心勃勃的超级项目的规模，该国已经委托顾问重新评估 170 公里长的线条城市 The Line 的可行性。The Line 是沙漠超级城市 Neom 的核心，高 500 米，长 170 公里，能容纳 900 万居民，城市被封闭在两座距离约 200 米的平行高墙之内，表面覆盖镜子。The Line 有三层，地面一层供行人使用，地下两层一层为基础设施一层为地下交通。它还有一条高铁，时速能达到 512 公里，从线条城市的一头到另一头只需要 20 分钟。整个城市完全使用可更新能源提供电力。建筑成本估计为 1000 亿到 2000 亿美元，项目的第一阶段计划在 2030 年完成。Neom 与 Red Sea 酒店和滑雪度假村是沙特王储 Mohammed bin Salman 的 Vision 2030 计划的旗舰项目，旨在转型沙特经济，减少对石油的依赖。但 Neom 项目陷入了困境，其项目负责人 Nadhmi al-Nasr 于 2024 年 11 月离职，另外两名外国高管也在年底离开。项目的进展也不顺利，预计到 2030 年只能完成 2.4 公里长度的建造工作，居民人数也将从预期的 150 万减少到 30 万。

在 Google 工作了 16 年的 Android 生态系统总裁 Sameer Samat 在接受采访时表示，该公司正计划将 ChromeOS 和 Android 合并为单一平台。Samat 在采访中谈论了 Android、Gemini、Galaxy AI 和 Android XR 等主题。对于合并 ChromeOS 和 Android，他没有透露多少信息，只是表示因为要统一 ChromeOS 和 Android， 他现在对用户如何使用笔记本电脑以及用笔记本电脑做什么非常感兴趣。