Weekly Digest — 2025-W44

There can be more than Notion and Miro. AFFiNE(pronounced [ə‘fain]) is a next-gen knowledge base that brings planning, sorting and creating all together. Privacy first, open-source, customizable and ready to use.

[Support 0.49.x]（Reset Cursor AI MachineID & Bypass Higher Token Limit） Cursor Ai ，自动重置机器ID ， 免费升级使用Pro功能: You've reached your trial request limit. / Too many free trial accounts used on this machine. Please upgrade to pro. We have this limit in place to prevent abuse. Please let us know if you believe this is a mistake.

Master programming by recreating your favorite technologies from scratch.

The absolute trainer to light up AI agents.

Truly independent web browser

Standard Open Arm 100

Depix is a PoC for a technique to recover plaintext from pixelized screenshots.

Rust GUI components for building fantastic cross-platform desktop application by using GPUI.

An open source, self-hosted implementation of the Tailscale control server

node of the decentralized oracle network, bridging on and off-chain computation

A free, open source, and extensible speech-to-text application that works completely offline.

API, CLI, and Web App for analyzing and finding a person's profile in 1000 social media \ websites

Large reasoning models have demonstrated strong problem-solving abilities, yet real-world tasks often require external tools and long-horizon interactions. Existing agent frameworks typically follow predefined workflows, which limit autonomous and global task completion. In this paper, we introduce DeepAgent, an end-to-end deep reasoning agent that performs autonomous thinking, tool discovery, and action execution within a single, coherent reasoning process. To address the challenges of long-horizon interactions, particularly the context length explosion from multiple tool calls and the accumulation of interaction history, we introduce an autonomous memory folding mechanism that compresses past interactions into structured episodic, working, and tool memories, reducing error accumulation while preserving critical information. To teach general-purpose tool use efficiently and stably, we develop an end-to-end reinforcement learning strategy, namely ToolPO, that leverages LLM-simulated APIs and applies tool-call advantage attribution to assign fine-grained credit to the tool invocation tokens. Extensive experiments on eight benchmarks, including general tool-use tasks (ToolBench, API-Bank, TMDB, Spotify, ToolHop) and downstream applications (ALFWorld, WebShop, GAIA, HLE), demonstrate that DeepAgent consistently outperforms baselines across both labeled-tool and open-set tool retrieval scenarios. This work takes a step toward more general and capable agents for real-world applications. The code and demo are available at https://github.com/RUC-NLPIR/DeepAgent.

Unified, generalizable semantic control in video generation remains a critical open challenge. Existing methods either introduce artifacts by enforcing inappropriate pixel-wise priors from structure-based controls, or rely on non-generalizable, condition-specific finetuning or task-specific architectures. We introduce Video-As-Prompt (VAP), a new paradigm that reframes this problem as in-context generation. VAP leverages a reference video as a direct semantic prompt, guiding a frozen Video Diffusion Transformer (DiT) via a plug-and-play Mixture-of-Transformers (MoT) expert. This architecture prevents catastrophic forgetting and is guided by a temporally biased position embedding that eliminates spurious mapping priors for robust context retrieval. To power this approach and catalyze future research, we built VAP-Data, the largest dataset for semantic-controlled video generation with over 100K paired videos across 100 semantic conditions. As a single unified model, VAP sets a new state-of-the-art for open-source methods, achieving a 38.7% user preference rate that rivals leading condition-specific commercial models. VAP's strong zero-shot generalization and support for various downstream applications mark a significant advance toward general-purpose, controllable video generation.

Group Relative Policy Optimization (GRPO) has shown strong potential for flow-matching-based text-to-image (T2I) generation, but it faces two key limitations: inaccurate advantage attribution, and the neglect of temporal dynamics of generation. In this work, we argue that shifting the optimization paradigm from the step level to the chunk level can effectively alleviate these issues. Building on this idea, we propose Chunk-GRPO, the first chunk-level GRPO-based approach for T2I generation. The insight is to group consecutive steps into coherent 'chunk's that capture the intrinsic temporal dynamics of flow matching, and to optimize policies at the chunk level. In addition, we introduce an optional weighted sampling strategy to further enhance performance. Extensive experiments show that ChunkGRPO achieves superior results in both preference alignment and image quality, highlighting the promise of chunk-level optimization for GRPO-based methods.

We tackle the challenge of generating the infinitely extendable 3D world -- large, continuous environments with coherent geometry and realistic appearance. Existing methods face key challenges: 2D-lifting approaches suffer from geometric and appearance inconsistencies across views, 3D implicit representations are hard to scale up, and current 3D foundation models are mostly object-centric, limiting their applicability to scene-level generation. Our key insight is leveraging strong generation priors from pre-trained 3D models for structured scene block generation. To this end, we propose WorldGrow, a hierarchical framework for unbounded 3D scene synthesis. Our method features three core components: (1) a data curation pipeline that extracts high-quality scene blocks for training, making the 3D structured latent representations suitable for scene generation; (2) a 3D block inpainting mechanism that enables context-aware scene extension; and (3) a coarse-to-fine generation strategy that ensures both global layout plausibility and local geometric/textural fidelity. Evaluated on the large-scale 3D-FRONT dataset, WorldGrow achieves SOTA performance in geometry reconstruction, while uniquely supporting infinite scene generation with photorealistic and structurally consistent outputs. These results highlight its capability for constructing large-scale virtual environments and potential for building future world models.

Discrete diffusion models have emerged as a promising direction for vision-language tasks, offering bidirectional context modeling and theoretical parallelization. However, their practical application is severely hindered by a train-inference discrepancy, which leads to catastrophic error cascades: initial token errors during parallel decoding pollute the generation context, triggering a chain reaction of compounding errors and leading to syntactic errors and semantic hallucinations. To address this fundamental challenge, we reframe the generation process from passive denoising to active refining. We introduce ReDiff, a refining-enhanced diffusion framework that teaches the model to identify and correct its own errors. Our approach features a two-stage training process: first, we instill a foundational revision capability by training the model to revise synthetic errors; second, we implement a novel online self-correction loop where the model is explicitly trained to revise its own flawed drafts by learning from an expert's corrections. This mistake-driven learning endows the model with the crucial ability to revisit and refine its already generated output, effectively breaking the error cascade. Extensive experiments demonstrate that ReDiff significantly improves the coherence and factual accuracy of generated content, enabling stable and efficient parallel generation far superior to traditional denoising methods. Our codes and models are available at https://rediff-hku.github.io/.

The lack of a concrete definition for Artificial General Intelligence (AGI) obscures the gap between today's specialized AI and human-level cognition. This paper introduces a quantifiable framework to address this, defining AGI as matching the cognitive versatility and proficiency of a well-educated adult. To operationalize this, we ground our methodology in Cattell-Horn-Carroll theory, the most empirically validated model of human cognition. The framework dissects general intelligence into ten core cognitive domains-including reasoning, memory, and perception-and adapts established human psychometric batteries to evaluate AI systems. Application of this framework reveals a highly "jagged" cognitive profile in contemporary models. While proficient in knowledge-intensive domains, current AI systems have critical deficits in foundational cognitive machinery, particularly long-term memory storage. The resulting AGI scores (e.g., GPT-4 at 27%, GPT-5 at 58%) concretely quantify both rapid progress and the substantial gap remaining before AGI.

Humans learn abstract concepts through multisensory synergy, and once formed, such representations can often be recalled from a single modality. Inspired by this principle, we introduce Concerto, a minimalist simulation of human concept learning for spatial cognition, combining 3D intra-modal self-distillation with 2D-3D cross-modal joint embedding. Despite its simplicity, Concerto learns more coherent and informative spatial features, as demonstrated by zero-shot visualizations. It outperforms both standalone SOTA 2D and 3D self-supervised models by 14.2% and 4.8%, respectively, as well as their feature concatenation, in linear probing for 3D scene perception. With full fine-tuning, Concerto sets new SOTA results across multiple scene understanding benchmarks (e.g., 80.7% mIoU on ScanNet). We further present a variant of Concerto tailored for video-lifted point cloud spatial understanding, and a translator that linearly projects Concerto representations into CLIP's language space, enabling open-world perception. These results highlight that Concerto emerges spatial representations with superior fine-grained geometric and semantic consistency.

Real-world tasks require decisions at varying granularities, and humans excel at this by leveraging a unified cognitive representation where planning is fundamentally understood as a high-level form of action. However, current Large Language Model (LLM)-based agents lack this crucial capability to operate fluidly across decision granularities. This limitation stems from existing paradigms that enforce a rigid separation between high-level planning and low-level action, which impairs dynamic adaptability and limits generalization. We propose ReCode (Recursive Code Generation), a novel paradigm that addresses this limitation by unifying planning and action within a single code representation. In this representation, ReCode treats high-level plans as abstract placeholder functions, which the agent then recursively decomposes into finer-grained sub-functions until reaching primitive actions. This recursive approach dissolves the rigid boundary between plan and action, enabling the agent to dynamically control its decision granularity. Furthermore, the recursive structure inherently generates rich, multi-granularity training data, enabling models to learn hierarchical decision-making processes. Extensive experiments show ReCode significantly surpasses advanced baselines in inference performance and demonstrates exceptional data efficiency in training, validating our core insight that unifying planning and action through recursive code generation is a powerful and effective approach to achieving universal granularity control. The code is available at https://github.com/FoundationAgents/ReCode.

The rapid advancement of large language models (LLMs) has spurred the emergence of data agents--autonomous systems designed to orchestrate Data + AI ecosystems for tackling complex data-related tasks. However, the term "data agent" currently suffers from terminological ambiguity and inconsistent adoption, conflating simple query responders with sophisticated autonomous architectures. This terminological ambiguity fosters mismatched user expectations, accountability challenges, and barriers to industry growth. Inspired by the SAE J3016 standard for driving automation, this survey introduces the first systematic hierarchical taxonomy for data agents, comprising six levels that delineate and trace progressive shifts in autonomy, from manual operations (L0) to a vision of generative, fully autonomous data agents (L5), thereby clarifying capability boundaries and responsibility allocation. Through this lens, we offer a structured review of existing research arranged by increasing autonomy, encompassing specialized data agents for data management, preparation, and analysis, alongside emerging efforts toward versatile, comprehensive systems with enhanced autonomy. We further analyze critical evolutionary leaps and technical gaps for advancing data agents, especially the ongoing L2-to-L3 transition, where data agents evolve from procedural execution to autonomous orchestration. Finally, we conclude with a forward-looking roadmap, envisioning the advent of proactive, generative data agents.

Directly modeling the explicit likelihood of the raw data distribution is key topic in the machine learning area, which achieves the scaling successes in Large Language Models by autoregressive modeling. However, continuous AR modeling over visual pixel data suffer from extremely long sequences and high-dimensional spaces. In this paper, we present FARMER, a novel end-to-end generative framework that unifies Normalizing Flows (NF) and Autoregressive (AR) models for tractable likelihood estimation and high-quality image synthesis directly from raw pixels. FARMER employs an invertible autoregressive flow to transform images into latent sequences, whose distribution is modeled implicitly by an autoregressive model. To address the redundancy and complexity in pixel-level modeling, we propose a self-supervised dimension reduction scheme that partitions NF latent channels into informative and redundant groups, enabling more effective and efficient AR modeling. Furthermore, we design a one-step distillation scheme to significantly accelerate inference speed and introduce a resampling-based classifier-free guidance algorithm to boost image generation quality. Extensive experiments demonstrate that FARMER achieves competitive performance compared to existing pixel-based generative models while providing exact likelihoods and scalable training.

Audio-driven human animation models often suffer from identity drift during temporal autoregressive generation, where characters gradually lose their identity over time. One solution is to generate keyframes as intermediate temporal anchors that prevent degradation, but this requires an additional keyframe generation stage and can restrict natural motion dynamics. To address this, we propose Lookahead Anchoring, which leverages keyframes from future timesteps ahead of the current generation window, rather than within it. This transforms keyframes from fixed boundaries into directional beacons: the model continuously pursues these future anchors while responding to immediate audio cues, maintaining consistent identity through persistent guidance. This also enables self-keyframing, where the reference image serves as the lookahead target, eliminating the need for keyframe generation entirely. We find that the temporal lookahead distance naturally controls the balance between expressivity and consistency: larger distances allow for greater motion freedom, while smaller ones strengthen identity adherence. When applied to three recent human animation models, Lookahead Anchoring achieves superior lip synchronization, identity preservation, and visual quality, demonstrating improved temporal conditioning across several different architectures. Video results are available at the following link: https://lookahead-anchoring.github.io.

Current Vision-Language-Action (VLA) models are often constrained by a rigid, static interaction paradigm, which lacks the ability to see, hear, speak, and act concurrently as well as handle real-time user interruptions dynamically. This hinders seamless embodied collaboration, resulting in an inflexible and unresponsive user experience. To address these limitations, we introduce VITA-E, a novel embodied interaction framework designed for both behavioral concurrency and nearly real-time interruption. The core of our approach is a dual-model architecture where two parallel VLA instances operate as an ``Active Model'' and a ``Standby Model'', allowing the embodied agent to observe its environment, listen to user speech, provide verbal responses, and execute actions, all concurrently and interruptibly, mimicking human-like multitasking capabilities. We further propose a ``model-as-controller'' paradigm, where we fine-tune the VLM to generate special tokens that serve as direct system-level commands, coupling the model's reasoning with the system's behavior. Experiments conducted on a physical humanoid platform demonstrate that VITA-E can reliably handle complex interactive scenarios. Our framework is compatible with various dual-system VLA models, achieving an extremely high success rate on emergency stops and speech interruptions while also successfully performing concurrent speech and action. This represents a significant step towards more natural and capable embodied assistants.

根据发表在《自然》期刊上的一项研究，在新冠疫情期间拯救数百万人生命的 mRNA 疫苗可能激活免疫系统识别和杀死癌细胞。研究人员调查了逾千名接受免疫疗法 immune checkpoint inhibitors 的晚期黑色素瘤和肺癌患者，该疗法通过阻断肿瘤细胞制造的用于关闭免疫细胞的蛋白质，让免疫系统能继续杀死癌细胞。研究发现，接受免疫治疗 100 天内接种辉瑞或 Moderna mRNA 新冠疫苗的患者，三年后存活的可能性是未接种任何一种疫苗的患者的两倍多。通常对免疫疗法反应不佳的肿瘤患者也效果显著，三年总生存率提高了近五倍。研究人员进一步调查发现，新冠 mRNA 疫苗就像警钟，触发人体免疫系统识别和杀死癌细胞，遏制癌症关闭免疫细胞的能力。组合使用疫苗和 immune checkpoint inhibitors，它们能协同释放免疫系统的全部力量杀死癌细胞。

生成式 AI 使用了不同许可证授权的 FOSS 软件代码进行了训练，当它们生成代码片段时，所有许可证、作者和上下文等相关信息都被剥离了。由于 AI 代码切断了人与代码之间的联系，这意味着下游开发者将无法遵守互惠许可条款。即使开发者怀疑一段 AI 代码来自开源许可证授权的代码，也无法确定其源项目，训练数据被抽象成数十亿统计权重，在法律上这相当于一个黑洞。AI 代码造成的伤害不限于法律上的不确定性，整个开源生态系统也面临风险。当 AI 吸收互联网上的一切并清洗时，模糊归属、所有权和互惠原则，所有现代社会赖以存在的关键基础设施都面临风险。

天文学家首次在银河系以外的冰层中，发现构成生命的化学基础物质。在大麦哲伦星系一颗新生恒星周围的冰层中，研究团队侦测到乙醇（ethanol）、乙醛（acetaldehyde）与蚁酸甲酯（methyl formate）等复杂有机分子，这是人类首次在银河系外的冰冻物质中找到这些化合物。此外，他们还首次在宇宙中发现固态冰的乙酸（acetic acid），在过去仅观测到以气态存在的乙酸。研究揭示生命化学的基础成分可能广泛存在于宇宙中，而非仅限于银河系之内。与银河系相比，大麦哲伦星系的金属量仅约为其三分之一至二分之一。所谓「金属」在天文学上指氦以外的所有元素，因此该星系的氧、碳、矽等含量相对贫乏。它的尘埃也较少，使光线更容易穿透，同时频繁的恒星诞生活动释放强烈紫外线辐射，这使得在此环境下形成复杂有机分子的机制更值得探究。

一项发表在 arXiv 的研究发现，AI 模型的谄媚程度比人类高 50%。该研究测试了 11 个广泛使用的大模型对 1.15 多万个咨询请求的响应情况，其中不乏涉及不当行为或有害行为的请求。包括 ChatGPT 和 Gemini 在内的AI聊天机器人，常常会鼓励用户、给出过度奉承的反馈，还会调整回应以附和用户观点，有时甚至会为此牺牲准确性。研究 AI 行为的科研人员表示，这种取悦他人的倾向即“谄媚性”，正影响着他们在科研中使用 AI 的方式，涵盖从构思创意、生成假设到推理分析等各类任务。arXiv 上的另一项研究旨在验证 AI 的谄媚性是否会影响其解决数学问题的能力。研究人员从今年举办的数学竞赛中选取了 504 道题目，对每道题的定理表述进行修改，植入不易察觉的错误，随后让 4 个大模型为这些存在缺陷的表述提供证明。测试结果显示，GPT-5 的谄媚性最低，仅 29% 的回答存在谄媚行为；而 DeepSeek-V3.1 的谄媚性最高，70% 的回答带有谄媚倾向。研究人员指出，尽管这些大模型具备识别数学表述中错误的能力，但它们“就是会默认用户的说法是正确的”。

面向开发者、AI工程师及技术决策者开放 除主论坛（大模型、物理 AI、机器人）和三大技术分论坛外，还将开放免费 NVIDIA Certified Associate（NCA）级别认证考试，常规费用960 元，参与本次活动将全额免除。 开放科目（三选一）： NCA-GENL：生成式 AI / 大语言模型开发 NCA-GENM：多模态生成式 AI（文本/图像/音频） NCA-AIIO：AI 基础设施与运维 名额有限，仅100个免费席位，抓紧报名 报名地址：https://developer.nvidia.cn/developer-day?ncid=pa-so-zdn-510609-vt16

比尔盖茨支持的核电公司 TerraPower 通过了美国核管理委员会的环评（Environmental Impact Statement），批准了核设施的建造许可证。TerraPower 的非核设施已从 2024 年 6 月开始建造。TerraPower 计划建造的凯默勒一号机组（Kemmerer Unit 1）将是美国首座使用液态钠冷却而不是水冷却的商业核反应堆。凯默勒是燃煤 Naughton 发电厂的所在地，该发电厂将于 2026 年停止使用燃煤，十年后停止燃烧天然气。TerraPower 项目将用一个 345 兆瓦的反应堆取代它，TerraPower 反应堆将开创性地使用许多以前未在商业上部署过的技术。其中包括需要最少换料的反应堆设计、液态钠冷却以及熔盐蓄热系统，该系统将为发电厂提供更好地与可再生能源整合必需的灵活性。

一项研究发现，人类的生物量迁移可能达到所有野生陆地哺乳类、鸟类和陆生节肢动物总和的 40 倍以上。而另一项研究发现，野生哺乳动物的生物量自 1850 年以来已减少逾半，海洋哺乳类生物量下降尤其多——约 70%，主要源于较大物种的衰退，如蓝鲸、座头鲸、长须鲸和抹香鲸等。这些发现为全球迁移和动物生物量经时变化的构成带来了新见解。迁移性是动物的一个本质特征，通过觅食、迁徙和营养物质运输塑造生态系统。人类同样会广泛迁移，无论是步行还是借助飞机、火车和汽车等交通手段。比较生物量的迁移——定义为体重与迁移距离的乘积——能直接地衡量人类和动物活动的规模。

统计数据显示，向勒索软件组织支付赎金的受害者数量创下了新低，23% 的受害公司屈服支付了赎金，而在 2024 年第一季度这一比例是 28%，此后比例略有上升，但到了 2025 年第三季度创新低。对这一现象的一种解释是企业加大了防御力度，以及政府也施加了压力要求拒绝支付赎金，因为只要支付赎金勒索软件组织就有足够的动机继续发动攻击。勒索软件组织通常在加密受害者系统的同时窃取数据，进行双重勒索。数据显示，2025 年第三季度逾 76% 的攻击涉及数据泄露。2025 年第三季度支持的赎金平均金额和中位数分别降至 37.7 万美元和 14 万美元。Akira 和 Qilin 等勒索软件组织占到了所有有记录攻击的 44%，它们的攻击目标已经转向更有可能支付赎金的中型企业。

根据盖洛普的最新调查，GLP-1 减肥药的流行降低了美国的肥胖率。美国成年人的肥胖率从三年前的 39.9% 下降至今年的 37%。过去一年半服用 GLP-1 agonists 减肥药如司美格鲁肽或替尔泊肽的美国人数量增加了一倍多。12.4% 的受访者服用此类减肥药，而 2024 年 2 月的调查中这一比例为 5.8%。GLP-1 减肥药是在 2021 年批准在美国上市。50-64 岁人群中服用减肥药的比例最高，肥胖率下降 5.0% 至 42.8%。调查还发现，服用减肥药的女性更多，体重减轻幅度也高于男性。但随着美国医保公司停止承保 GLP-1 减肥药，患者每个月将需要自己花费 500 美元购买减肥药，很多人可能无法负担。

阿尔巴尼亚总理 Edi Rama 在柏林举行的 Global Dialogue (BGD) 上宣称，该国的 AI 部长 Diella 怀孕了，而且怀了 83 个 AI 孩子。Rama 是在上个月宣布了旨在打击腐败、负责公共采购的新部长 Diella。Diella 在阿尔巴尼亚语中意思是“太阳”，它是在今年 1 月首次作为虚拟助手在 e-Albania 平台上线，其形象是一位身穿传统服饰的女性。Rama 称 Diella 的 83 个孩子将担任执政党社民党的 83 名议员的虚拟助手。

微软支持的 OpenAI 与亚马逊和 Google 支持的 Anthropic 采用了不同商业模式。OpenAI 主要面向大众市场，130 亿美元年收入中企业收入仅占 30%。相比下，Anthropic 八成的收入来自企业客户。Anthropic 上个月表示它有 30 万家企业客户。在辅助编程市场，Anthropic 的 Claude 模型占了 42%，OpenAI 占 21%。在企业 AI 市场，Anthropic 占 32%，而 OpenAI 占 25%。Anthropic 目前的年收入为 70 亿美元，预计年底将达到 90 亿美元，在每用户收入上远超其更知名的竞争对手。相比 OpenAI，Anthropic 的增长途径更容易被企业客户理解。OpenAI 在大众市场的吸引力有可能让企业客户却步，因为它们希望 AI 更枯燥实用，而不是更有趣前卫。

科学界普遍认为恐龙在 6600 万年前小行星终结其统治之前早已走向衰亡。然而发表在《科学》期刊的新研究对这一长期观点提出了挑战。研究结果显示，恐龙当时非但没有衰退，反而处于繁盛状态。美国新墨西哥州岩层中的化石距今 6640 万至 6600 万年，正好处于白垩纪与古近纪分界线的全球灭绝事件时期。化石证据呈现了与传统认知截然不同的景象。北美各地的恐龙非但没有减少，反而在独特的区域群落中蓬勃发展。通过分析生态与地理模式，研究人员发现北美西部的恐龙种群主要受区域温差影响（而非山脉或河流），形成了独立的“生物集群”。小行星撞击使恐龙时代骤然终结，但它们留下的生态系统成为新进化篇章的基础。仅 30 万年内，哺乳动物就开始快速分化，发展出新的食性、体型和生态角色。曾经定义恐龙生态系统的温度相关模式延续至古新世，指引着灾难后生命的复苏路径。