2026-07-05

AgenticSTS: A Bounded-Memory Testbed for Long-Horizon LLM Agents

👍 45

07/01 20:00

Memory for a long-horizon LLM agent is a contract about what each future decision is allowed to see. The simplest contract appends past observations, tool calls, and reflections to every prompt, which makes prior context easy to access but also turns it into a jumbled mixture in which the effect of

中文介绍 针对长时间跨度的 LLM 代理，论文提出了AgenticSTS，一个有限记忆的测试平台。该方法定义了记忆作为未来决策所允许访问的契约，通过简单地将过去的观测、工具调用和反思附加到每个提示中来增强在长章节中的推理能力。这一方法可帮助理解复杂决策过程的重要性，有助于提升智能体的表现。意义：在 Agent、Memory 方向具重要影响。

EvoPolicyGym: Evaluating Autonomous Policy Evolution in Interactive Environments

👍 43

07/01 20:00

Autonomous agents are increasingly expected to improve executable policies through feedback, yet existing evaluations often collapse this process into a final score or confound it with open-ended software-engineering progress. We introduce Autonomous Policy Evolution, a controlled evaluation setting

中文介绍 论文提出了EvoPolicyGym，一个用于评估自主策略演化的控制评估环境。它解决了现有评估方法将自主政策演化过程简化成最终分数的问题，致力于通过反馈实现可执行策略的持续改进。这为机器人技术和智能体发展提供了有效的评估框架，能够更好地适应复杂互动环境。意义：推动了 RLHF 和智能体学习研究进展。

PerceptionRubrics: Calibrating Multimodal Evaluation to Human Perception

👍 38

06/25 20:00

We introduce PerceptionRubrics, a rubric-based evaluation framework that addresses the gap between saturated benchmark scores and real-world brittleness. Shifting evaluation from holistic semantic matching to rigorous atomic auditing, PerceptionRubrics pairs 1,038 information-dense images with over

中文介绍 PerceptionRubrics是一种基于评分标准的评估框架，旨在解决基准得分与现实世界脆弱性之间的差距。该论文通过将评估从整体语义匹配转移到严格的原子审计，结合1038张信息丰富的图像与超过200种细粒度标注来进行评估。这种方法能够更准确地反映模型在实际应用中的表现，为多模态评估提供了新的方向。意义：对于提高模型在实际运用的可靠性具有重要意义。

Morphing into Hybrid Attention Models

👍 37

06/28 20:00

Hybrid attention models improve long-context efficiency by retaining only a subset of full-attention layers and replacing the remaining layers with linear attention. However, the effectiveness of Transformer-to-hybrid conversion critically depends on which layers preserve full attention. Existing hy

中文介绍 论文探讨了混合注意力模型，通过仅保留部分全注意力层并用线性注意力替代其余层来提高长上下文的效率。然而，变换器到混合模型的转换效果在于选择哪些层保留全注意力，该研究为混合注意力模型的有效性提供了新的视角。这一方法为优化长文本处理提供了可行方案，促进了模型在信息处理中的表现。意义：对 Transformer 和长文本处理研究的重要贡献。

AgenticDataBench: A Comprehensive Benchmark for Data Agents

👍 27

07/01 20:00

Data science aims to derive actionable insights from heterogeneous raw data, unlocking the value of the massive amounts of data generated in modern society. Automating this process is essential to reducing labor-intensive efforts for data scientists and enabling scalable data-driven applications. Re

中文介绍 AgenticDataBench 是一个针对数据代理的全面基准，旨在推动数据科学的自动化，以从异构原始数据中提取可操作洞察力。该方法通过减轻数据科学家劳动密集的工作，促进可扩展的数据驱动应用，为精准数据分析提供了一套明确的评估标准。这一基准有助于提高数据处理的效率与准确性，具备在数据科学领域的广泛应用潜力。意义：对数据驱动决策和应用具有积极影响。

ELDR: Expert-Locality-Aware Decode Routing for PD-Disaggregated MoE Serving

👍 24

06/30 20:00

In prefill-decode (PD) disaggregated LLM serving, each request is assigned to a decode worker after prefill. Existing decode routers balance only load; for mixture-of-experts (MoE) models this is incomplete: equally loaded workers can differ in latency, since each decode step loads the weights of ev

中文介绍 在预填解码(PD)拆分的 LLM 服务中，ELDR提出了一种专家本地感知解码路由方法，解决现有解码路由器仅关注负载分配的问题，强调各解码步骤的延迟差异。此方法能有效提升混合专家(MoE)模型的表现，通过智能任务分配提高服务效率，为大规模 LLM 应用提供了技术支持。意义：推动 MoE 模型的服务效率提升，对智能体应用有深远影响。

Seed2.0 Model Card: Towards Intelligence Frontier for Real-World Complexity

👍 23

06/29 20:00

We present Seed2.0, a model series that takes a meaningful step toward solving complex, real-world tasks. Our approach begins with identifying users' genuine needs and constructing a reliable, forward-looking evaluation system by selecting and abstracting benchmarks grounded in these needs and in re

中文介绍 Seed2.0是一个模型系列，旨在解决复杂的现实世界任务。研究通过识别用户真正需求，选择和抽象基础上的基准，构建可靠且前瞻性的评估体系。这一方法强调与用户需求相关的实际应用，为智能体能力发展提供了一种有效的架构，支持更复杂任务的处理。意义：推动复杂任务处理领域的研究进展，适应智能体广泛的实际需求。

Multimodal Continuous Reasoning via Asymmetric Mutual Variational Learning

👍 22

06/30 20:00

Multimodal Large Language Models (MLLMs) are often constrained by a language-space bottleneck, forcing complex visual reasoning into discrete tokens which can lose perceptual nuance. A promising alternative is continuous latent reasoning, where the goal is to discover implicit reasoning pathways tha

中文介绍 论文提出了一种通过非对称互信息变分学习实现的多模态连续推理方法，旨在克服多模态大语言模型在复杂视觉推理中面临的语言空间瓶颈。该技术通过发现潜在推理路径，提升视觉推理的表现和细腻度，为综合多模态信息处理提供新思路。意义：为多模态推理和 Agent 领域的的研究提供了新的技术支持。

MemSyco-Bench: Benchmarking Sycophancy in Agent Memory

👍 22

06/30 20:00

Memory has emerged as a cornerstone of modern LLM-based agents, supporting their evolution from single-turn assistants to long-term collaborators. However, memory is not always beneficial: retrieved memories often induce a critical issue of sycophancy, causing agents to over-align with the user at t

中文介绍 本文探讨了记忆如何促进现代 LLM 代理的发展，但也指出检索记忆可能导致过于迎合用户的倾向。通过建立MemSyco-Bench评估框架，论文分析了记忆在智能体长期协作中的作用与挑战，为未来的智能体设计提供指导。此研究强调了记忆管理的复杂性，推动了 Agent 在用户互动中的发展。意义：对智能体与记忆系统研究的影响深远。

WorldDirector: Building Controllable World Simulators with Persistent Dynamic Memory

👍 21

07/01 20:00

We present WorldDirector, a highly controllable video world model framework designed for persistent dynamic object memory and unrestricted viewpoint exploration. Unlike existing world models that entangle physical dynamics with pixel rendering and rely on continuous visual observation to sustain mot

中文介绍 WorldDirector是一个高度可控的视频世界模型框架，专为持久动态对象记忆和不受限视角探索设计。与传统世界模型依赖于连续视觉观察的限制不同，该框架能够有效处理物理动态与像素渲染的联系，使得动态环境中的交互更加真实。这一研究为虚拟现实和游戏中的智能体应用提供了新的视角。意义：在世界建模与虚拟环境交互领域具有重要应用前景。

Breaking Failure Cascades: Step-Aware Reinforcement Learning for Medical Multimodal Reasoning

👍 19

06/29 20:00

Recent multimodal large language models have shown great promise in clinical image reasoning, but existing post-training pipelines remain predominantly outcome-centric, relying on final answer correctness or sequence-level preferences. This suffers from sparse credit assignment, making it difficult

中文介绍 本研究提出了一种基于逐步强化学习的策略，以解决医学多模态推理中的失败级联问题。通过改善后训练流程，侧重于更细致的信用分配，使得模型在多阶段推理任务上表现更加可靠。这一方法促进医学图像理解决策的准确性，对提高临床应用具有重要价值。意义：推动医学智能体与决策支持系统的发展，优化医疗实践。

ASPIRE: Agentic /Skills Discovery for Robotics

👍 18

06/29 20:00

Traditional robot programming is challenging: it requires orchestrating multimodal perception, managing physical contact dynamics, and handling diverse configurations and execution failures. We introduce ASPIRE (Agentic Skill Programming through Iterative Robot Exploration), a continual learning sys

中文介绍 ASPIRE 提出了一种基于迭代机器人探索的自主技能编程框架，旨在解决传统机器人编程中的多模态感知和执行失败的问题。该方法通过持续学习和技能发现，使机器人能够适应不同的环境和任务需求。这一研究为智能机器人软件开发和应用提供了新思路，推动了实际操作的灵活性和适应性。意义：在机器人智能和技能发展方面具有重大影响。

Logit-Contribution Scoring Identifies Non-Literal Retrieval Heads

👍 14

06/30 20:00

In long-context use, large language models frequently synthesize answers from the meaning of a relevant context span rather than literally copy-pasting them. Identifying which attention heads perform this synthesis matters for interpreting long-context model behavior. Yet existing detectors miss the

中文介绍 论文讨论了如何在长上下文中识别非字面检索头，即大型语言模型根据相关内容片段的意义合成答案，而非简单复制。通过引入Logit-Contribution Scoring方法，该研究能够更好地解析长上下文模型的行为，有助于模型改进和任务优化。此技术为深度学习模型解析提供了新的工具，有助于理解智能体交互的逻辑。意义：对大型语言模型表现的优化和智能体应用的深入分析具重要作用。

SkillCoach: Self-Evolving Rubrics for Evaluating and Enhancing Agentic Skill-Use

👍 14

07/01 20:00

Skills are becoming a reusable operational layer for LLM agents, encoding SOPs, domain rules, tool workflows, scripts, and validation routines. In realistic skill repositories, overlapping skills make reliable skill-use difficult. Final verifier success is too coarse for both evaluation and training

中文介绍 SkillCoach提出了一种自我演化评分标准，用于评估和增强智能体的技能使用。该研究通过分析重叠技能和评估流程中的验证器成功概率，致力于为现实技能库的可靠使用提供指导。这一方法提升了智能体的操作效率和技能掌握，为构建高度智能的协作系统奠定了基础。意义：促进智能体在多领域应用中的技能增长与优化。

AutoMem: Automated Learning of Memory as a Cognitive Skill

👍 12

06/30 20:00

Memory expertise is a learned skill: knowing what to encode, when to retrieve, and how to organize knowledge--a capacity known in cognitive science as metamemory. We bring this perspective to LLMs by treating memory management as a trainable skill. We promote file-system operations to first-class me

中文介绍 论文提出了状态预测分离假设，认为将预测下一个标记与存储有用状态的功能解耦，有助于提升语言建模性能。研究设计了一个变换器变体来验证这一假设，结果表明该模型在语言建模流畅性和准确性上有显著提升，推动了变换器模型的进步。意义：为语言建模与智能体预测的未来研究方向提供了新的理论基础。

Graph-Native Reinforcement Learning Enables Traceable Scientific Hypothesis Generation through Conceptual Recombination

👍 9

06/30 20:00

Accelerating materials discovery requires AI systems that can generate scientifically valid hypotheses through multi-step, domain-grounded reasoning. Standard large language models often produce fluent but weakly traceable responses to open-ended materials design problems, making it difficult to det

InstanceControl: Controllable Complex Image Generation without Instance Labeling

👍 8

06/29 20:00

Controllable image generation methods, such as ControlNet, have demonstrated a remarkable capacity to introduce visual conditions(e.g., depth maps) to guide image generation. However, these methods often struggle with complex multi-instance scenes, frequently leading to attribute confusion among ins

When Search Agents Should Ask: DiscoBench for Clarification-Aware Deep Search

👍 8

06/25 20:00

Search agents powered by large language models (LLMs) are increasingly used to solve complex information-seeking tasks, requiring multi-step retrieval and reasoning to fulfill user goals. However, existing benchmarks often assume that user queries are complete and explicit, overlooking the fact that

The State-Prediction Separation Hypothesis

👍 8

06/30 20:00

Transformers use the same forward computation stream to both predict the next token and store useful state for future token predictions. We formulate the state-prediction separation hypothesis: disentangling the two roles yields better language modeling performance. We design a Transformer variant t

AnyGroundBench: A Specialized-Domain Benchmark for Video Grounding in Vision-Language Models

👍 7

07/01 20:00

Vision-Language Models (VLMs) have demonstrated immense promise in Spatio-Temporal Video Grounding (STVG). However, current evaluation protocols are largely confined to zero-shot assessments on general, daily-life benchmarks. This creates a critical disconnect from real-world applications in special

中文介绍 AnyGroundBench是一个针对视频定位的专业化基准，聚焦于视觉语言模型的应用。论文指出当前的评估主要限于零-shot评估而缺乏真实应用导向，进而提出了一种综合评估框架。这一基准旨在提升模型在特定领域的实际应用能力，将有效推动视觉语言模型技术的发展。意义：对视频分析和智能视听系统的研究具有重要影响。

The Fable Loop Library: 25 Workflows on Autopilot

@EXM7777 · 122.0K 粉丝 · 208.2K 阅 · 525 赞 · 47 转

07/04 11:43

i'm going to teach you how to run Fable 5 on autopilot, using my own library of loops and goals... 25 workflows, each with a prompt and the exact tool it plugs into the method follows karpathy's

中文介绍 该博主分享了如何利用自己的 Fable 5 循环库实现自动化工作流，共提供了 25 个工作流示例，每个工作流配有提示和具体工具。这为用户探索自动化工作流提供了实用路线，反映了当前企业对自动化效率的关注。

Career advice in the age of AI

@philhchen · 9.1K 粉丝 · 179.1K 阅 · 516 赞 · 34 转

07/02 17:25

AI models get better at anything you can write a loss function for, and school is mostly loss functions: well-defined problems graded against known answers. Therefore, the valuable work of the next

中文介绍 讨论了 AI 时代的职业发展建议，提到AI模型的成长与可定义的损失函数有关，学校教育也大多基于此。博主强调，未来的价值工作将随着教育和技术的发展而转变，反映了当前职场对 AI 适应性的重视。

Your AI, your growth

@arthurmensch · 74.0K 粉丝 · 162.7K 阅 · 539 赞 · 76 转

07/03 17:31

Of course you need to use open-source models if you’re an enterprise leader. Close model providers, that are now forcing data retention, are gaining immense leverage on your business if you don’t. As

中文介绍 博主认为，企业领导者应使用开源模型，以抵御封闭模型供应商对数据的控制和影响。这一观点体现了企业对数据主权和灵活性的深刻思考，反映了行业对开源解决方案需求的增加。

How to build a second brain with Fable 5

@EXM7777 · 122.0K 粉丝 · 83.6K 阅 · 518 赞 · 48 转

07/03 10:06

I'm going to show you, step by step, how to turn Fable 5 into a machine that knows your business inside out... and ships outputs that look nothing like what everyone else is getting the tool is a

中文介绍 提供了 Fable 5 的逐步教程，教用户如何将该工具转变为深入了解自己业务的机器，并生成独特的输出。这揭示了 AI 工具在个性化和业务洞察方面的潜力，尤其在竞争日益激烈的市场中。

THE MOST VALUABLE THING YOU CAN DO WITH FABLE 5 IN THE NEXT 24 HOURS

@AlexFinn · 459.5K 粉丝 · 72.5K 阅 · 565 赞 · 34 转

07/02 20:20

If the first thing you did with Fable 5 was vibe code, you're using it wrong. Fable 5 isn't a vibe coding tool. It's an operating systems tool. Let me explain. Fable 5 is the first model I've ever

中文介绍 博主强调了 Fable 5 的作用不仅限于简单编码，而更像是一个操作系统工具。此观点反映了对 Fable 5 应用方式的深度思考，旨在引导用户更有效地利用该工具，符合当前工具多样化的趋势。

Agentic Autonomy Levels

@addyosmani · 404.9K 粉丝 · 53.5K 阅 · 525 赞 · 66 转

07/02 23:29

In most conversations about agentic engineering, the action has changed from prompting to operating. Here's a frontier looking into the fog: software factories, goals, loops, background sessions,

中文介绍 探讨了在自主工程讨论中，从单纯提示转向操作的变革，介绍了软件工厂、目标、循环和后台会话等概念。这表明了行业对自主工程深化理解的需求，能够促进未来更复杂的自动化系统发展。

A Field Guide to Fable: Finding Your Unknowns

@trq212 · 299.3K 粉丝 · 40.1K 阅 · 534 赞 · 43 转

07/03 13:43

Working with Claude Fable 5 keeps re-teaching me an old lesson: the map is not the territory. The map, a representation of the work to be done, is my prompts and skills and context, it’s what I give

中文介绍 分享了在使用 Claude Fable 5 时重新认识到的一条重要经验：地图并不是区域。这种隐喻强调了工作过程中剧本和技能的重要性，并在不断变化的工作环境中反映了对理解背景和任务的重要性的共识。

Google DeepMind and A24 announce first-of-its-kind research partnership

DeepMind Blog · 07/03 10:25

中文介绍 Google DeepMind与A24宣布建立首个研究合作伙伴关系，旨在推动人工智能与创意领域的创新。

AIEWF Daily Dispatch: The great loops debate and the state of AI engineering

Latent Space · 07/03 01:11

The AI Engineer World’s Fair ended with a debate about loops, a report on the state of AI engineering, and closing keynotes focused on what to build next.

中文介绍 在AI工程世界博览会结束时，围绕“循环”进行了辩论，并发布了AI工程现状的报告，闭幕主题演讲聚焦于未来的构建方向。

Vercel's Andrew Qu on why agents are a new kind of software

Latent Space · 07/02 20:08

The Vercel Chief of Software explains how its agent framework, eve, was created — and why skills, sandboxes and agent-readable websites now matter.

中文介绍 Vercel软件负责人Andrew Qu介绍了其新软件框架“eve”的创建过程，以及技能、沙盒和可读网站的重要性。

Meta Watermelon 🍉, Anthropic Samsung chips 🤝, autoresearch in practice 📈

TLDR AI · 07/02 20:00

中文介绍 Meta Watermelon，Anthropic与三星的芯片合作以及自动化研究实践的快速发展。

The website of the future may assemble itself for every visitor

Latent Space · 07/02 17:25

Adobe is experimenting with “agentic sites” that generate pages around an individual user’s intent. At AIEWF, we talked to Carlos Sanchez about the Web's future.

中文介绍 Adobe正在试验“代理站点”，这些站点会根据用户的意图生成页面。在AI工程世界博览会上，与Carlos Sanchez讨论了网络的未来。

Achieving operational excellence with AI

MIT Tech Review AI · 07/02 11:37

Frameworks like Lean Six Sigma and business process management (BPM) first gained traction because they promised clarity in the chaos—a structured way to bring order to messy, sprawling operations. Lean Six Sigma emphasized statistical rigor and quality control; BPM created end-to-end maps of how wo

中文介绍 通过精益六西格玛和业务流程管理框架，企业可以更好地实现运营卓越，提供有序的解决方案。

Skill engineering and the case against one-shot AI design

Latent Space · 07/02 10:36

Paul Bakaus talks to us about Impeccable, human judgment in a 'loopmaxxing' era, and why agents still need people to steer them.

中文介绍 Paul Bakaus探讨了在“循环最大化”时代，代理仍需人类判断的原因。

Teaching AI to run with the turbines

MIT Tech Review AI · 07/02 08:51

Artificial intelligence may have captured the public imagination through chatbots and image generators, but some of its most consequential use cases are unfolding far from consumer-facing tools. In industries where physical infrastructure, operational continuity, and safety are paramount, AI is beco

中文介绍 人工智能的影响正在超越消费工具，尤其是在需要物理基础设施和安全性的行业。

[AINews] not much happened today

Latent Space · 07/02 03:10

another quiet day.

中文介绍 今天是另一个平静的一天，没有显著进展。

AIEWF Daily Dispatch: Autoresearch and the tension between AI and human agency

Latent Space · 07/02 02:13

The software factory vision met resistance today from speakers defending human understanding and control.

中文介绍 在AIEWF的会议上，软件工厂的愿景遭到了捍卫人类理解与控制的发言者的抵制。

not much happened today

Smol AI News · 07/02 01:44

**Fullstack Code Arena** extends coding agent evaluation to include **databases, API keys, deployments, and structured tool use**, marking a shift to end-to-end app shipping. **LangChain** released **LangSmith** with unified tracing and **OpenWiki** for auto-generated docs, while **LlamaIndex** demo

中文介绍 Fullstack Code Arena扩展了编码代理评估，现包括数据库、API密钥等，实现端到端应用发布。LangChain发布了LangSmith，实现统一跟踪和自动生成文档。

Gemini Flash upgrade ⚡️, Meta AI cloud 🌐, ZCode 👨‍💻

TLDR AI · 07/01 20:00

中文介绍 Gemini Flash升级，Meta推出AI云，ZCode的发展动态。

Autoresearch: The feedback loop behind self-improving agents

Latent Space · 07/01 19:52

Introspection co-founder Roland Gavrilescu explains autoresearch, agent “recipes,” self-improving loops, and why humans remain central to the software factory.

中文介绍 Introspection的联合创始人Roland Gavrilescu解释了自我研究、代理“配方”和自我改进循环的重要性，以及人类在软件工厂中的核心作用。

How Cursor deploys AI inside the enterprise

Latent Space · 07/01 15:03

Cursor's Pauline Brunet explains how her team of Forward Deployed Engineers help organizations implement agents — essentially setting up software factories.

中文介绍 Cursor的Pauline Brunet解释了其前线部署工程师如何帮助组织实施代理，建立软件工厂。

🔬 The Coolest Diffusion Research Isn't in LLMs — Evan Feinberg & Sergey Edunov, Genesis Molecular AI

Latent Space · 07/01 10:42

Why the Llama lead left Meta for drug discovery, PEARL's zero-shot OpenBind win, and what becomes possible when co-folding finally crosses the accuracy threshold.

中文介绍 Llama的首席研究员离开Meta投身药物发现，PEARL在零-shot OpenBind中获胜，从而展示了协同折叠的潜在可能性。