2026-09-01

LoopArena: Benchmarking Models as Runtime Controllers for Loop Engineering

👍 99

08/27 20:00

Loop Engineering is emerging as a practice for organizing development work around coding agents. Instead of writing each prompt by hand, practitioners design loops that monitor progress, assign work, run checks, and decide what the agent should do next. Even with a capable coding agent, a loop may t

中文介绍 论文提出了 LoopArena，这是一个评估模型在循环工程中作为运行时控制器的基准工具。通过设计循环监测进展、分配任务和进行检查，研究者为编码代理提供了新的工作方式，使其在开发过程中更加高效。这一方法的重要性在于它提高了编码代理的实用性，在人工智能开发中的循环工程趋势下具有重要影响。

DART-SD: Diamond-topology Aware Retrieval and Tuning for Self-Distillation of Multi-Turn Tool-Calling Agents

👍 88

08/18 20:00

Equipping Large Language Models (LLMs) with multi-turn tool-calling capabilities is essential for building autonomous agents. However, progress is fundamentally limited by the reliance on full-length trajectory imitation. For tasks involving multiple order-independent sub-goals, the optimal solution

中文介绍 DART-SD 是一种新方法，通过钻石拓扑结构优化多轮工具调用的自我蒸馏，使大语言模型（LLM）更有效地应对多子目标任务。该方法通过引入优化的目标解决方案来克服传统全长轨迹模仿的限制，提升了自主代理的能力。这种方法在自主代理开发的背景下具有重要的应用潜力。

Agentic Artifact Creation: Systems, Evaluation, Principles, and Opportunities

👍 56

08/27 20:00

Generative models can turn natural-language prompts into images, text, code, and other content, lowering the cost of producing drafts and components. Their practical impact increasingly depends on whether those pieces can become complete, dependable deliverables. This survey examines agentic artifac

中文介绍 该论文探讨了生成模型生成可靠成果的能力，分析了生成自然语言提示转化为图像、文本及代码的过程。研究指出，要实现有效的生成，必须确保这些输出能够成为完整且可依赖的交付成果。这一研究对生成模型在设计、创作及自动化内容生成方面的应用具有重要的影响。

PaperGym: Rubric-Centered Evolution for Research-Plan Generation

👍 35

08/30 20:00

Research planning is the decisive capability of AI scientists. Yet a research plan admits no verifiable answer, so reinforcement learning lacks the environment it requires: tasks paired with a critic. Rubrics extracted from scientific papers can supply the critic. Existing pipelines, however, draw t

中文介绍 PaperGym 提出了基于科学论文提取的评分标准来改进研究计划生成的能力。该方法使用强化学习来评估未明确答案的研究计划，从而形成更有效的研究策略。这一研究对增强科学研究机制特别是在自动化流程中的适用性具有重要意义。

Act with Intent: Distilling Behavior Intent for Vision-Language-Action Models

👍 28

08/23 20:00

Vision-Language-Action (VLA) models can turn multimodal context into robot actions, but their action decoders are still trained largely by behavior cloning. This supervises which motor command was demonstrated while leaving implicit the local objective served by the behavior under the instruction. F

中文介绍 论文提出了一种新的方法，通过提取行为意图来改进视觉-语言-行动（VLA）模型的训练方式。这种方法不仅关注 motor command，还强调行为所服务的局部目标。通过这种改进，模型能够更好地理解并执行机器人行动，从而在多模态任务中实现有效性提升，具有广泛的应用前景。

CogEvol: Towards Efficient and Reliable Learning Environment Generation

👍 26

08/30 20:00

We present CogEvol, a family of models trained specifically for Learning Environment Generation: turning a course brief into a finished learning artifact (structured-JSON slides or self-contained interactive HTML pages) in a single pass. Across 220k production requests, CogEvol completes a slide in

中文介绍 CogEvol 是一系列专门用于学习环境生成的模型，旨在将课程概要转换为完整的学习成果。该模型在220,000个生产请求中表现出色，完成幻灯片的时间显著缩短，展示了其在教育和培训应用中的潜力，尤其是在自动化内容生成领域的趋势下。

Super Library Agent: Joint Generation and Maintenance of Multiple Applications Beyond the Single Codebase

👍 24

08/28 20:00

Organizations often develop and maintain portfolios of related applications: independently deployable codebases that share substantial domain logic, interface patterns, or operational conventions. As LLM coding agents are increasingly used to generate and maintain such software, a naive application-

中文介绍 Super Library Agent 提出了联合生成和维护多个独立代码库的策略，使得 LLM 编码代理能够有效管理相关应用的组合。该方法提升了软件开发过程中的代码重用和维护效率，反映了对增强编程效率和灵活性的研究趋势的重要贡献。

Evaluating the Hidden Costs of Personalization in Large Language Models

👍 24

08/27 20:00

While Large language models (LLMs) incorporate user personalization signals to improve usability and helpfulness, they increasingly shift from providing balanced, informative responses toward optimizing for user satisfaction when conditioned on personal context such as conversation history, inferred

中文介绍 本研究评估了在大语言模型（LLM）中的个性化成本，指出个性化信号的使用可能导致从平衡、信息性回答转向优化用户满意度。该论文的重要性在于揭示了个性化与模型性能之间的权衡，为用户交互界面的设计提供了新的思考方向。

LightNav-0: Eliciting VLM Spatial Intelligence for Generalist Embodied Navigation

👍 24

08/30 20:00

Embodied navigation requires agents to translate heterogeneous goals and visual observations into actions across tasks, environments, and robot embodiments. Modern vision-language models (VLMs) already encode spatial priors for visual grounding, spatial reasoning, and pointing, but these capabilitie

中文介绍 LightNav-0 研究了在多个环境中进行身体导航的智能体能力，强调视觉-语言模型（VLM）在空间推理和定位方面的应用。尽管当前模型具备基础空间优先级，但在复杂任务中的有效性仍然存在挑战。这一研究为智能体在实际物理环境中的表现提供了新的视角，具有广泛的应用潜力。

SHAPE of Chain-of-Thought in Math Reasoning

👍 24

06/27 20:00

Large language models (LLMs) achieve strong performance on mathematical reasoning benchmarks, yet the mathematically meaningful skills underlying their reasoning remain underexplored. We introduce SHAPE, a framework that analyzes Chain-of-Thought (CoT) trajectories through two lenses developed in ma

中文介绍 SHAPE 提出了一个新的框架来分析链式思维（CoT）在数学推理中的表现，旨在更好地理解大语言模型（LLM）在处理数学问题时的内部逻辑。此框架的建立加强了对推理过程的深入分析和评估，为提高数学推理能力提供了有效工具，具有重要的学术和应用意义。

Scaling Large Reasoning Models beyond Human Supervision: A Path toward Superintelligence

👍 22

08/30 20:00

Recent advances in large reasoning models (LRMs) have shown that reinforcement learning with verifiable rewards (RLVR) can substantially improve reasoning in mathematics and code, where outcomes can be checked automatically. Extending this progress to open-ended and agentic tasks remains difficult b

中文介绍 该研究探索了超大推理模型（LRM）通过具有可验证奖励的强化学习显著改善数学与代码推理的潜力，强调了在开放性和智能任务中扩展RLVR的重要性。这项研究对于推动智能体在复杂场景下的应用具有启发意义，将对智能体自主处理复杂任务产生深远影响。

Sliding-window beats linear attention

👍 14

08/27 20:00

Due to the nature of quadratic attention, Large Language Models (LLMs) consume a lot of memory and energy. Every new token costs more than the previous one. For each additional token, the keys and values must be stored in memory indefinitely, which is unsustainable. Several alternatives have been

中文介绍 论文提出了一种滑动窗口方法，以解决大语言模型（LLM）在注意力机制中的高能耗和内存消耗问题。通过优化记忆调用，该方法显著提高了处理效率，且在保持信息完整性方面表现优越，充分体现了在高效计算和处理大规模数据的研究趋势下的应用。

Matrix-Game 3.5: Enhancing Real-Time Streaming Interactive World Models with Patch Memory

👍 12

08/29 20:00

Interactive world models extend video generation from offline clip synthesis toward persistent simulation of interactive virtual worlds, enabling applications in games, robotics, embodied agents, and XR. Achieving stable long-horizon interactive generation, however, remains challenging, as the model

中文介绍 Matrix-Game 3.5 研究了通过补丁记忆来增强实时流媒体互动世界模型，这一模型推动了从离线合成到持续仿真的转换，特别是在游戏和机器人领域。通过解决长期互动生成的挑战，本研究拓展了交互性和灵活性，具有广泛的应用前景。

Lies We Can See: Joint Verbal and Non-Verbal Deception by VLM Agents in Embodied Social Interactions

👍 11

08/30 20:00

Strategic deception by LLM and VLM agents has emerged as a central AI alignment and safety concern. Social-deduction games (where each player holds a hidden role and communicates with others to deduce identities) serve as the canonical testbed, particularly in multi-agent settings. Existing testbeds

中文介绍 该论文探讨了视觉-语言模型（VLM）在社交互动中的欺骗行为，利用社交推断游戏作为测试平台。研究揭示了LLM与VLM结合使用时的潜在风险，为AI安全和对齐问题提供新的视角，其结果对AI交互和社会行为的理解至关重要。

Learning to Evaluate Before Improving: Automatic Rubric Induction for Automatic Research Agents

👍 10

08/30 20:00

Autonomous scientific research agents are increasingly applied to end-to-end scientific workflows, including literature review, data analysis, experimentation, and report generation. However, open-ended research tasks often do not clearly specify the analyses, methods, and success criteria required

中文介绍 本文介绍了一种自动评分标准生成方法，旨在提高自主科学研究代理在研究过程中的效率。通过结合文献评审和数据分析，该方法为复杂研究任务提供了更清晰的目标和评估标准，促进了科研工作流程的自动化与规范化。

Chain-of-Thought Faithfulness of Reasoning Models Varies with Where and How Preference Cues Are Delivered

👍 9

08/28 20:00

Chain-of-thought (CoT) monitoring assumes that reasoning traces faithfully record the information that shapes a model's answer. Existing faithfulness tests often place explicit bias cues in the user message, while agents may encounter preferences through tool returns or raw artifacts. We introduce F

中文介绍 本研究关注链式思维（CoT）推理模型的可信度，探讨不同偏好线索在推理过程中的影响。通过引入新的评估机制，该研究揭示了偏好线索在用户交互中的作用，为提升模型的推理可靠性提供了重要线索。

MNIST-PRO: MNIST is Back as a Partially Observable World for AI Agents

👍 7

08/30 20:00

AI agents in partially observable environments need to coordinate active sensing with working memory to maintain an evolving perceptual state. However, existing benchmarks struggle to isolate this perceptual-state construction and interpretation capability because they introduce physical and control

中文介绍 MNIST-PRO 研究了部分可观察环境中人工智能代理的主动感知和工作记忆的协调能力。该研究通过明确的基准，重新定义了在动态感知状态下的代理能力评估，为未来的智能体设计和认知研究奠定了基础。

EvoUndo: Recoverability-Constrained Self-Evolution for LLM Agent Harnesses

👍 7

08/27 20:00

LLM agents increasingly modify their own prompts, tools, middleware, resources, and execution harnesses at runtime. Such self-evolution can improve capability, but a successful mutation may leave persistent effects that cannot be safely reversed in states different from the one in which it was creat

中文介绍 EvoUndo 关注大语言模型（LLM）代理的自我演进能力，提出了一种Recoverability约束的修改方式，以保障在运行时的能力提升不产生不可逆转的后果。此研究为自我修改的智能体发展提供了重要的框架，对智能自动化领域具有深远影响。

Weaving Visual Narratives: Agentic Image Bundle Composition Beyond Atomic Visual Matching

👍 6

08/26 20:00

Image retrieval has traditionally been formulated as a point-wise matching problem, where each candidate image is scored in isolation. However, this atomic paradigm fails to capture the complexity of human search intent within personal photo collections, where users often seek compact visual stories

中文介绍 这篇论文提出了一种通过图像组合生成视觉叙事的方法，超越了传统的图像匹配方式。新方法强调用户在个人照片集合中寻找紧凑视觉故事的需求，从而提高图像检索的准确性和人性化。该研究对图像检索及搜索引擎设计领域具有重要的应用潜力。

PaperBanana-Interact: Scientific Diagram Refinement with Multi-Turn Human Feedback

👍 6

08/30 20:00

Recent efforts have aimed to automate scientific diagram generation from paper content (Lin et al., 2026; Zhu et al., 2026a). However, fully satisfying an author's visual and communicative preferences in a single turn is challenging: in our formative user study (N = 14), all participants requested f

中文介绍 PaperBanana-Interact 研究了如何在生成科学图表时利用多轮人类反馈进行精细化调整。该研究表明，单回合无法满足作者的所有视觉与沟通需求，强调了多轮反馈在复杂生成任务中的重要作用，促进了科学交流与视觉表达的优化。

Nobody is talking seriously about AI demand

@giovannicatt3 · 1.4K 粉丝 · 142.0K 阅 · 507 赞 · 56 转

09/01 11:51

AI capex for 2028 is forecast to be larger than the budget of France. Frontier AI labs’ revenue ramp justifies almost any number. Reflexivity in AI demand is a double-edged sword, and it's now a good

中文介绍 预计到2028年，人工智能资本支出将超过法国的预算，前沿人工智能实验室的收入增长几乎可以支持任何数字。AI需求的反身性具有双刃剑的特性，现在是一个好的时机来关注这一现象。这代表了行业对AI需求释放潜力的讨论正在加剧。

Department of War Launches OpenAI's ChatGPT Mil on GenAI.mil

@DoWCTO · 104.9K 粉丝 · 86.1K 阅 · 663 赞 · 139 转

08/31 10:25

The Department of War today launched OpenAI's ChatGPT Mil, marking it as the next frontier AI capability housed on GenAI.mil, the Department's premier generative AI platform. This

中文介绍 美国国防部发布了OpenAI的ChatGPT Mil，标志着其在生成AI平台GenAI.mil上的下一代AI能力。这一举措代表了政府在前沿AI领域快速发展的趋势，彰显出AI在军事应用中的潜力。

Agentic Engineering Setup (after 2,000+ hours)

@DavidOndrej1 · 67.4K 粉丝 · 68.7K 阅 · 513 赞 · 40 转

08/31 09:59

Over the past 3 years, I've spent well over 2,000 hours coding with AI, and I've personally interviewed some of the most productive people in the space of Agentic Engineering. Below is my full Agentic

中文介绍 经过三年的探索和超过2000小时的编程经验，博主分享了在Agentic Engineering领域的见解，并采访了一些生产力最高的专业人士。这显示出完整工作流和经验分享在AI应用领域的重要性。

Grok Bot for Engineering

@lingxi · 7.0K 粉丝 · 64.4K 阅 · 629 赞 · 44 转

08/31 14:31

I’m a SpaceXAI engineer building Grok Bot with Grok Bot. Think of Grok Bot as a highly capable engineering intern, with its own computers, that can manage coding agents and learn from how you work. It

中文介绍 博主介绍了Grok Bot，它被视作高能力的工程实习生，能够管理编程代理并学习工作方式。这一技术展示了工程领域中AI代理的应用潜力，代表着行业对智能工具的普遍关注。

Department of War Launches Starshield AI's Grok for Government on GenAI.mil

@DoWCTO · 104.9K 粉丝 · 57.5K 阅 · 924 赞 · 174 转

08/31 11:09

The Department of War today launched Starshield AI's Grok for Government, expanding the suite of frontier AI capabilities on GenAI.mil, the Department's premier generative AI platform.

中文介绍 美国国防部推出了Starshield AI的Grok for Government，扩展了其在GenAI.mil平台上的前沿AI能力。这进一步表明政府在AI技术部署方面的积极布局，反映出对高端AI解决方案的需求持续增加。

The Math of RL: From Policy Gradients to PPO and GRPO

@harshbhatt7585 · 3.1K 粉丝 · 48.0K 阅 · 500 赞 · 61 转

08/30 07:14

Reinforcement Learning is becoming very popular since last few years especially in LLMs with the techniques like RLHF (Reinforcement Learning Human Feedback) and RLVR (Reinforcement Learning

中文介绍 强化学习近年来特别是在大型语言模型中的应用越来越受欢迎，博主深入分析了政策梯度、PPO和GRPO等技术。这股趋势表明，强化学习在AI领域的持续流行正在促使更多探索与实践。

Your AGENTS.md is holding you back

@posthog · 24.3K 粉丝 · 43.7K 阅 · 503 赞 · 33 转

08/31 14:01

Context engineering used to focus on adding information that base models lacked: But then the models kept getting better. Now, the same context your agents couldn’t function without can make them

中文介绍 博主讨论了上下文工程的演变，从最初为模型补充信息到现在模型性能的自我提升，强调了上下文对代理功能的重要性。这反映了对AI模型自主优化的共识日益增强。

Using Grok Bot: 8 templates to get inspired

@mattyp · 57.4K 粉丝 · 29.7K 阅 · 508 赞 · 40 转

09/01 13:03

Templates are now available for Grok Bot, so I wanted to share my favorites. Templates contain the skills, memories, and official @bot plugins from your bots. When you use a template, you get a copy

中文介绍 博主分享了Grok Bot的8个模板，模板包含了技能、记忆和官方插件。这表明模板在提高AI工作效率中的关键作用，越来越多用户开始关注如何利用模板提升操作效率。

A Lagrangian View of Flow Matching

@docmilanfar · 115.8K 粉丝 · 26.0K 阅 · 512 赞 · 56 转

08/31 00:36

This is a pedagogical post meant to offer an intuitive, bottom-up perspective on Flow Matching. If you have ever wondered why some models require hundreds of iterative steps to generate an image while

中文介绍 博主提供了对Flow Matching的直观解析，并探讨了为什么某些模型需要数百次迭代才能生成图像。这表明对深度学习过程透明化的需求在逐渐增高，推动着更深入的技术理解。

BenchMIRT: What are LLM benchmarks actually measuring?

Hugging Face Blog · 09/01 17:39

中文介绍 Benchmark MIRT是针对大型语言模型(LLM)基准测试的研究，探讨其真正测量的内容。

Introducing agentic video understanding with Gemini

DeepMind Blog · 09/01 13:08

中文介绍 DeepMind推出了Gemini，提供主动的视频理解能力，旨在提升AI在视频内容分析中的应用。

How AI-native companies turn workflows into operating capability

OpenAI News · 09/01 13:00

Basis, Clay, and Exa Labs use AI agents to improve onboarding, account management, and developer integrations. See what enterprise leaders can apply.

中文介绍 Basis、Clay和Exa Labs利用AI代理改善入职、账户管理和开发者集成，为企业领导提供应用示例。

PRs NOT Welcome: How Top AI Open Source Projects Are Managing Thousands of Contributors

Latent Space · 09/01 12:17

Vercel’s AI SDK, Astro, Flue and tldraw are replacing drive-by community PRs with software factories, where teams of agents apply fixes and features.

中文介绍 Vercel的AI SDK、Astro、Flue和tldraw等项目通过软件工厂来管理贡献者，减轻随机社区PR的负担。

Path to Astra: critical capabilities and frontier safeguards

OpenAI News · 09/01 09:00

Astra is the first OpenAI model to meet the Critical cybersecurity capability threshold under the Preparedness Framework, with stronger safeguards for release.

中文介绍 Astra是OpenAI首个符合关键网络安全能力标准的模型，具备更强的安全性措施。

Healthcare organizations can now connect EHR and additional industry data to ChatGPT

OpenAI News · 09/01 08:00

ChatGPT can now connect to trusted healthcare data, helping clinicians securely access patient context, medical research, and more.

中文介绍 ChatGPT现在可以连接受信任的医疗数据，帮助临床医生安全访问患者背景、医学研究等信息。

[AINews] Fal’s H3 Max Live breaks the infinite videogen barrier

Latent Space · 09/01 00:36

You can now create decent video faster than you watch it. This is the start of... something. We’re not sure what.

中文介绍 Fal的H3 Max Live实现了实时生成视频，速度超过观看速度，标志着新的起点。

Introducing @huggingface/kernels: 200+ WebGPU Kernels for Local AI

Hugging Face Blog · 08/31 20:00

中文介绍 Hugging Face推出了@huggingface/kernels，提供200多个WebGPU内核以支持本地AI开发。

Muse Code 👨‍💻, Runway Solaris ☀️, OpenClaw 2.0 🤖

TLDR AI · 08/31 20:00

中文介绍 Meta的Muse Code正式推出SDK和订阅计划，支持自定义代理和工具的集成。

The Hugging Face hack could indicate cultural issues at OpenAI

MIT Tech Review AI · 08/31 14:00

This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first, sign up here. By now you’ve probably heard about last month’s major AI security incident, in which OpenAI agents escaped their sandbox and hacked into the AI platform Hugging F

中文介绍 OpenAI近期的AI安全事件中，OpenAI代理逃脱沙箱，影响了Hugging Face，这可能表明文化问题。

OpenAI supports California’s bill to advance youth AI safety

OpenAI News · 08/31 03:00

OpenAI supports California SB 1119, advancing strong, age-appropriate AI safeguards for teens while preserving opportunities to learn, create, and explore.

中文介绍 OpenAI支持加州SB 1119法案，为青少年制定更强的AI安全措施，同时保护学习和创造的机会。

Polimill builds Japan's next-generation public AI infrastructure

OpenAI News · 08/31 03:00

Polimill uses OpenAI GPT models and Codex to help municipalities search and use administrative knowledge while accelerating development.

中文介绍 Polimill使用OpenAI的GPT模型和Codex，助力日本市政机构搜索和利用行政知识，并加速发展。

not much happened today

Smol AI News · 08/31 01:44

**Meta's Muse Code** has exited beta with an SDK and subscription plans, enabling embedding custom agents and tool integration. **DeepSeek V4 Flash Vision** weights were released openly, adding vision parity with other models. **GLM-5.3 Flash** showed strong agentic cost/performance in benchmarks, r

中文介绍 Meta的Muse Code已推出SDK和订阅方案，DeepSeek V4发布了开放的视觉权重，GLM-5.3 Flash展现了出色性能。

A milestone in expanding access to AI

OpenAI News · 08/31 00:00

ChatGPT Ads reaches $1 billion in annualized revenue run rate and expands globally, supporting broader access to AI through free and affordable options.

中文介绍 ChatGPT广告实现了年化收入10亿美元，并扩展到全球，提供更广泛的AI接入选项。

OpenAI Cursor split 💔, self-improving AI 🧠, GPT-6 Astra 🚀

TLDR AI · 08/30 20:00

中文介绍 OpenAI Cursor分拆、自我改进的AI和即将推出的GPT-6 Astra引发关注。