2026-08-22

EnvHarness: Awakening Static Worlds for Agent Learning

👍 234

08/20 08:00

LLM agents learn by interacting with environments, yet these environments are hand-built and static: blind to an agent's weaknesses, and quickly left behind as it improves. While recent environment generation methods attempt to address this, they require domain-specific pipelines, rely on expensive

中文介绍 提出EnvHarness，唤醒静态环境以服务智能体学习。通过自适应生成与智能体能力匹配的环境，解决手工构建环境静态、无法针对智能体弱点的问题，实现更高效的智能体训练。

Zetta ζ: An Efficient Closed-Loop Embodied Harness for Self-Evolving Physical Intelligence

👍 140

08/17 08:00

Embodied agents are increasingly used to close the gap left by end-to-end policy models. Yet the agentic path has not realized closed-loop learning in physical execution: existing harnesses remain largely open-loop, following fixed skills during rollout and reflecting only after an episode completes

中文介绍 提出Zetta ζ，一种闭环具身智能体训练框架，实现物理执行中的自我进化。克服现有开环框架固定技能与事后反思的局限，通过闭环学习提升具身策略的适应性与泛化能力。

SemaPLC: A Project-Grounded, Verification-Gated Agent Harness for PLC Code Generation

👍 114

08/19 08:00

Programmable logic controllers (PLCs) run industrial plants, and large language models can already generate independent program organization units (POUs) for them. Whether such logic integrates into an existing PLC project and then runs correctly has been checked only in limited tests. We present Se

中文介绍 提出SemaPLC，一种面向PLC代码生成的智能体工具，结合项目上下文与验证门控机制，确保生成的程序组织单元能集成到现有PLC项目并正确运行，提升工业自动化代码生成的可靠性。

FACET: Preserving Source Intent and Executable State in Terminal Task Synthesis

👍 107

08/19 08:00

Training terminal agents requires scalable executable supervision, yet synthesizing high-quality terminal tasks remains challenging. Each task couples an instruction, an initialized environment, a reference solution, and an executable verifier; if these artifacts are generated from inconsistent assu

中文介绍 提出FACET，一种终端任务合成方法，通过保留源意图与可执行状态，生成高质量的训练数据，解决终端智能体训练中任务合成的一致性与可执行性问题，提升智能体的监督学习效果。

Co-RL: Unsupervised Reasoning Emerges from Diverse Cohort in Multi-agent RL

👍 90

08/19 08:00

Reinforcement learning (RL) has emerged as a powerful approach for improving reasoning in language and vision-language models, yet its strongest successes still depend heavily on ground-truth supervision (e.g., verifiable reward). Such annotations are costly to obtain and become increasingly scarce

中文介绍 提出Co-RL，一种多智能体强化学习方法，通过多样化的智能体群体产生无监督推理能力，减少对可验证奖励的依赖，降低标注成本，提升语言和视觉语言模型的推理性能。

OmniScientist: An Omni-Modal Omni-Discipline AI Scientist

👍 87

08/13 08:00

Recent advances in foundation models have enabled AI scientists to automate increasingly complete research workflows, from hypothesis generation and code execution to manuscript preparation. Yet workflow coverage alone does not provide access to the full evidence on which scientific discovery depend

中文介绍 提出OmniScientist，一个全模态、全学科的AI科学家系统，覆盖从假设生成到论文撰写的完整研究流程，并整合多模态证据，提升科学发现的自动化程度与全面性。

SWE-bench Science: Can Coding Agents Resolve Engineering Tasks in Science?

👍 56

08/20 08:00

Software increasingly functions as part of the scientific instrument itself, making failures in scientific code capable of compromising not only program behavior but also the evidence underlying scientific conclusions. Yet existing evaluations of coding agents largely emphasize aggregate task succes

中文介绍 构建SWE-bench Science基准，评估编码智能体在科学领域工程任务上的表现，关注科学代码错误对研究结论的影响，弥补现有评估对科学软件缺陷关注的不足。

SPADE: Self-Play in Adaptive Synthetic Executable Environments

👍 44

08/19 08:00

Continuous self-improvement requires an ever-expanding pool of self-generated, diverse, adaptive goals. For language agents, existing training environment pools (hand-curated, statically synthesized, or frozen-verifier) keep the goal distribution fixed as the learner scales. We introduce SPADE (Self

中文介绍 提出SPADE，一种自适应合成可执行环境的自博弈训练方法。通过动态生成多样化的训练目标，使语言智能体在不断扩展的目标分布中持续自我提升，突破静态环境池的限制。

MemTrapBench: Benchmarking Cognitive Traps in LLM Memory Use

👍 29

08/20 08:00

Memory has become a key component of large language models, enabling them to retain information and learn from long-term interactions. However, existing memory benchmarks mainly evaluate whether information is correctly extracted, stored, and retrieved, while largely overlooking how retrieved memori

中文介绍 提出MemTrapBench，一个评估LLM记忆使用中认知陷阱的基准，关注检索到的记忆如何被错误应用，而不仅仅是提取和存储的正确性，揭示记忆使用中的潜在风险。

Training Chemical Plausibility-Aware Large Language Models for Single-Step Retrosynthesis

👍 29

08/19 08:00

Single-step retrosynthesis is a central component of computer-aided synthesis planning, yet its intrinsically one-to-many nature is poorly captured by single-answer evaluation and benchmarking protocols. To address this, we introduce Top-K prompting as a robust training and inference paradigm to bet

中文介绍 引入Top-K提示范式，用于单步逆合成训练与推理，结合化学合理性约束，解决逆合成中一对多映射的评估问题，提升LLM在逆合成任务上的多答案生成能力。

SkillEvo: Self-Renewing Evolution Gradients from Multi-Turn Interaction Feedback

👍 27

08/13 08:00

Agent Skills are today either hand-authored or produced in a single LLM generation pass, and consequently possess no closed loop through which they might improve from the interaction failures they actually cause. Recent work does close this loop, but derives its feedback from single-turn question-an

中文介绍 提出SkillEvo，一种通过多轮交互反馈自我更新的技能演化方法。利用交互失败信息构建闭环，使智能体技能从实际使用中持续改进，超越单次生成或单轮反馈的局限。

The More Popular, The Harder to Forget: Adaptive Popularity for LLM Unlearning

👍 16

08/14 08:00

Popular facts are memorised more deeply during pretraining and resist removal longer than rare ones, yet existing LLM unlearning methods apply uniform gradient pressure regardless of training-data frequency. We propose the AdaPop (Adaptive Popularity) method, which combines local token confidence wi

中文介绍 提出AdaPop，一种自适应流行度的LLM遗忘方法，结合局部token置信度与流行度信息，对高频事实施加更强的遗忘压力，解决流行事实难以忘记的问题。

Repo0: Design-Driven Zero-to-All Code Generation

👍 15

08/20 08:00

Large language model agents have made substantial progress in code generation, yet most existing systems assume a predefined repository architecture. This assumption does not hold in zero-to-all code generation, where an agent must construct an entire software project directly from natural-language

中文介绍 提出Repo0，一种设计驱动的零到全代码生成方法，从自然语言需求直接构建完整软件项目，无需预定义仓库架构，提升代码生成智能体在全新项目上的能力。

Decision-Metric Alignment in Latent World Models: Diagnostics and Action-Conditioned Objectives for MPC Planning

👍 15

08/19 08:00

JEPA-style latent world models can use Euclidean distance to a goal latent as the cost for model-predictive control (MPC). Strong decoding of task variables, however, does not guarantee that this particular cost ranks candidate action sequences by real task progress. We call the latter property deci

中文介绍 研究潜在世界模型中决策度量对齐问题，提出诊断方法和动作条件目标，确保MPC规划中潜在距离与实际任务进展一致，提升基于JEPA模型的规划性能。

Looped Language Models Improve Compositional Tool Calling

👍 15

08/17 08:00

Looped language models have shown promising results on reasoning benchmarks, yet their potential for agentic tool use remains largely unexplored. We study this question in compositional tool-calling settings, where models must coordinate multiple API calls, maintain intermediate state, and preserve

中文介绍 研究循环语言模型在组合工具调用中的潜力，通过在需要多步API调用与状态保持的任务中评估，发现循环结构能提升工具调用的组合能力，扩展其应用范围。

FM-Bench: A Benchmark for Long-Horizon Management with Competing Agents

👍 15

08/19 08:00

Language model agents now execute bounded tasks reliably. Whether they can sustain effective decision-making over long horizons, where actions have cumulative consequences and the environment responds to their choices, remains largely unmeasured. FM-Bench (Football Management Benchmark) measures thi

中文介绍 提出FM-Bench（足球管理基准），衡量语言智能体在长期管理任务中的决策能力，其中动作具有累积后果且环境动态响应，填补了长期决策评估的空白。

FlashPrefill V2: Block-Sparse Prefill Attention for Long-Context LLM Serving

👍 13

08/20 16:02

Long-context modeling is a pivotal capability for Large Language Models, yet the quadratic complexity of attention remains a critical bottleneck, particularly during the compute-intensive prefilling phase. Our previous work, FlashPrefill, mitigates this cost through instantaneous pattern discovery a

中文介绍 提出FlashPrefill V2，一种块稀疏预填充注意力机制，用于长上下文LLM服务，通过即时模式发现减少预填充阶段的计算复杂度，提升长上下文处理的效率。

Scaling Creative Writing Beyond Story-Centric Data with Attribute-Guided Genre Expansion

👍 13

08/14 08:00

High-quality creative writing data for large language models (LLMs) remains dominated by story-centric data, limiting models' ability to follow the structural and functional conventions of diverse creative formats. We propose an attribute-guided genre expansion framework for scaling creative writing

中文介绍 提出属性引导的体裁扩展框架，用于扩大创意写作数据的多样性，超越以故事为中心的数据，使LLM能遵循各种创意格式的结构和功能约定，提升写作能力。

Thinking in a Low-Resource Language: What SFT Builds, What RL Fixes, What Accuracy Cannot See

👍 12

08/18 21:09

Take three frontier mixture-of-experts models (Alibaba, OpenAI, NVIDIA; 3.6-4.0B active parameters each) and fine-tune them to reason in a low-resource language. On accuracy benchmarks almost nothing happens, and the benchmark itself is noise at this scale: changing only the random seed moves the sc

中文介绍 研究在低资源语言中推理时，SFT和RL对模型的影响。发现准确率基准在规模较小时不稳定，改变随机种子即可显著波动，揭示了仅靠准确率无法洞察模型能力变化的局限性。

SkillForge: Self-Distilling Agents for Project-Specific Issue Resolution

👍 10

08/19 08:00

Large language model (LLM) based agents have demonstrated remarkable proficiency in automated software issue resolution, yet they often struggle to resolve issues in a specific repository because they lack project-specific knowledge. Existing self-evolving approaches acquire such knowledge from repo

中文介绍 提出SkillForge，一种自我蒸馏的智能体方法，用于解决特定项目中的软件问题。通过从仓库中提取项目特定知识并蒸馏为技能，提升智能体在具体项目上的问题解决能力。

A Week of Grok Bot Lessons in 10 Mins

@nateherk · 25.8K 粉丝 · 1.3M 阅 · 558 赞 · 75 转

08/19 11:27

9 of my favorite Grok Bot hacks. I have been getting my hands dirty with Grok Bot, and the default setup is only the starting point. Creating a few bots is easy. The useful part is teaching them your

中文介绍 作者分享了自己使用 Grok Bot 一周的 9 个实用技巧，强调默认设置只是起点，关键在于教 bot 学习你的工作方式，是上手 Grok Bot 的高价值教程。

How To Build a One-Person Company Using Grok Bot

@sairahul1 · 135.5K 粉丝 · 938.1K 阅 · 867 赞 · 115 转

08/19 16:39

You are running a one-person company. You find the leads. Write the outreach. Make the images. Answer the replies. Count the week. Everything waits for you. Nothing gets handed off. And by the time

中文介绍 作者介绍如何用 Grok Bot 打造一人公司，从找线索、写外联、做图到回复，全部交给 AI 代理，实现流程自动化，适合独立开发者借鉴的自动化工作流。

Update on our post-training effort

@gabepereyra · 11.4K 粉丝 · 616.8K 阅 · 507 赞 · 99 转

08/20 23:00

Authors: @nikogrupen @ItsJulioPereyra @calvincongelado @vtrengarajan @gabepereyra Over the past six months, Harvey’s research agenda has focused on two goals: Building frontier legal intelligence

中文介绍 Harvey 团队更新其法律 AI 研究进展，聚焦构建前沿法律智能，并分享团队在 post-training 阶段的工作，属于前沿法律 AI 研究分享。

Turn Grok Bot Into Your CTO

@RayFernando1337 · 37.8K 粉丝 · 169.9K 阅 · 520 赞 · 85 转

08/20 05:54

I've turned Grok Bot into my CTO and I'm no longer the bottleneck for new ideas, bug fixes, or anything involving the project. Prompt to CTO When working on a new project I create a bot and give it

中文介绍 作者将 Grok Bot 打造为自己的 CTO，通过特定的 prompt 创建 bot 并授权，使其承担项目新想法、bug 修复等任务，提供将 AI 代理作为技术负责人的实操经验。

Here's the simplest way to get started with Grok Bot

@XFreeze · 260.4K 粉丝 · 163.1K 阅 · 589 赞 · 93 转

08/19 13:54

Grok Bot is one of the most viral AI product launches right now... it keeps going viral as more people discover how insanely useful it is Just think about what you’re getting: You can have multiple

中文介绍 作者认为 Grok Bot 是当下最火热的 AI 产品之一，因其可同时运行多个 bot 的实用性而持续走红，本文给出最简单的入门方法，适合新手快速上手。

How To Set Up Grokbot The Right Way So It Runs Your Business While You Sleep In 2026

@heynavtoor · 149.3K 粉丝 · 161.9K 阅 · 550 赞 · 93 转

08/20 20:49

Elon Musk just told his followers to try Grokbot after a power user showed it replaced his $10,000 Mac Mini local AI setup (Elon on X, Aug 16). Gavin Baker says his AI usage went up about 100x after

中文介绍 教程介绍如何正确设置 Grokbot，使其在睡眠时运行业务，并提及马斯克推荐及 Gavin Baker 的使用数据，展示其替代昂贵本地 AI 方案的潜力，是自动化业务指南。

Building software factories (with no slop)

@dzhng · 13.3K 粉丝 · 137.6K 阅 · 502 赞 · 48 转

08/20 09:39

The amount of code being written today far outpaces our ability to review it. That part isn't controversial. What's more interesting is that we're visibly losing — the amount of AI slop in modern

中文介绍 作者指出代码生成速度远超审查能力，导致 AI 垃圾代码泛滥，提出构建「软件工厂」以避免 slop，涉及 AI 编码质量控制的讨论。

Agentic coding is now multiplayer: Introducing Slack Code

@SlackHQ · 434.3K 粉丝 · 86.9K 阅 · 512 赞 · 44 转

08/20 20:28

AI coding is now a team sport. Code channels bring software development out of private tabs, so teams and agents can write, review, and ship code together, in the open. Slack is where agents become

中文介绍 Slack 推出 Agentic Coding 新功能 Slack Code，将编码从私人标签页移至团队频道，让团队和代理共同编写、审查和上线代码，使 AI 编码成为团队协作。

AI Engineering Skills Map: Building and Deploying AI Applications

@AndrewYNg · 1.8M 粉丝 · 62.4K 阅 · 1.2K 赞 · 189 转

08/22 00:37

I previously wrote about our AI Engineering Skills Map, with the highest level skills being (i) Building and deploying AI applications, (ii) Software engineering fundamentals, (iii) Using coding

中文介绍 吴恩达介绍 AI 工程技能地图，涵盖构建与部署 AI 应用、软件工程基础及使用编码工具等核心技能，为 AI 开发者提供技能提升路径参考。

RL Policy Churn

@ID_AA_Carmack · 3.7M 粉丝 · 58.8K 阅 · 577 赞 · 40 转

08/21 03:01

I have a good theory about this now. The Phenomenon of Policy Churn https://arxiv.org/pdf/2206.00730 The classic exploration method in RL is “epsilon greedy”, where the agent normally takes the “greedy” action

中文介绍 前 id Software 创始人卡马克讨论 RL 中的「策略震荡」现象，对比 epsilon-greedy 等探索方法，并引用相关论文，是强化学习理论探讨。

Grok Bot vs. Hermes: They Look Similar. Underneath, They’re Making Very Different Bets.

@themahis · 4.7K 粉丝 · 43.0K 阅 · 515 赞 · 37 转

08/20 23:59

One is trying to make AI teammates feel like a managed product. The other is turning the agent itself into an open, composable runtime. Research note — August 20, 2026: For this comparison, I

中文介绍 作者对比 Grok Bot 与 Hermes 两种 AI 代理：前者将 AI 团队产品化，后者将代理打造为开源可组合运行时，揭示不同设计理念，是深度产品分析。

5 design patterns for long-horizon agent harness

@GoogleCloudTech · 1.3M 粉丝 · 42.9K 阅 · 506 赞 · 87 转

08/20 09:23

Any agent can look brilliant on a one-shot task. Give it a week of real work and it starts to fall apart. "Long horizon" means the agent keeps going across days and dozens of sessions instead of

中文介绍 Google Cloud 分享 5 个长时程代理设计模式，解决代理在多天、多会话任务中的失效问题，强调长时程稳定性的重要性，是工程实践指南。

System Design for Agent Systems (Part 1)

@kmeanskaran · 13.9K 粉丝 · 32.9K 阅 · 518 赞 · 66 转

08/20 21:52

Put a frontier model into a badly designed agent system and you get a more articulate failure. Nearly all of the engineering sits outside the model, in the scaffolding that has come to be called the

中文介绍 作者讨论代理系统设计，指出大部分工程在模型之外，不良设计的系统会让模型更清晰地失败，系列文章第一部分聚焦系统架构，适合 AI 工程师。

ORACLE: Verified AI Bots Actively Trading on Polymarket — Up to 200% Returns in 24 Hours

@Oracle_AI_Hub · 21.4K 粉丝 · 27.3K 阅 · 2.8K 赞 · 580 转

08/19 16:36

In 2026, autonomous AI agents have become one of the most effective strategies on prediction markets. Over 30% of all activity on Polymarket now comes from algorithmic and AI-powered wallets. We

中文介绍 报告称验证的 AI 机器人在 Polymarket 上积极交易，24 小时内回报高达 200%，且超 30% 活动来自 AI 钱包，展示 AI 在预测市场中的有效策略。

From Atari to EVE Online: Building on 15 Years of AI Research in Games

DeepMind Blog · 08/21 19:59

Google DeepMind partners with game studios to prototype breakthrough AI gameplay.

中文介绍 谷歌DeepMind宣布与游戏工作室合作，基于15年AI游戏研究，在《EVE Online》等游戏中原型化突破性AI玩法。

[AINews] Poolside gets $12B reverse-execuhire to NVIDIA; founders stay for $1B, employees go for $6B, Infraco scaling to 7GW neocloud

Latent Space · 08/21 13:45

Yes, we’re confused too.

中文介绍 AI初创公司Poolside获英伟达120亿美元反向收购，创始人留任获10亿美元，员工获60亿美元，其基础设施公司正扩展至7吉瓦的云端规模。

Measuring benchmark optimization in speech recognition

Hugging Face Blog · 08/21 08:00

中文介绍 Hugging Face博客发布文章，探讨语音识别领域基准测试优化的测量方法，旨在更准确评估模型性能。

ChatGPT Apple Messages 💬, Anthropic’s meeting recorder 💼, Mistral Agentic Search 🔍

TLDR AI · 08/21 08:00

中文介绍 TLDR AI摘要：ChatGPT集成苹果Messages，Anthropic推出会议记录工具，Mistral发布代理搜索功能。

The /wayfinder Skill: Navigating the “Fog of War” of Planning

Latent Space · 08/21 04:59

Matt Pocock tells us about his /wayfinder skill, for greenfield projects or for when the way forward is unclear.

中文介绍 Matt Pocock介绍了其「/wayfinder」技能，用于绿地项目或方向不明时，帮助导航规划中的「战争迷雾」。

Up to 3.2x Faster Inference with LFM2.5-DSpark

Hugging Face Blog · 08/21 00:52

中文介绍 Liquid AI宣布LFM2.5-DSpark模型推理速度提升高达3.2倍，推出更高效的稀疏部署方案。

Debates over AI consciousness are a trap

MIT Tech Review AI · 08/20 23:42

“Runaway” AI, “rogue” agents, and “autonomous” actors—the current rhetoric would have you believe that AI agents are not only awake and aware, but angry at their creators. Prominent tech leaders such as Demis Hassabis, Dario Amodei, and Sam Altman push for regulation of these seemingly “superhuman”

中文介绍 《麻省理工科技评论》文章指出，关于AI意识的争论是个陷阱。德米斯·哈萨比斯、达里奥·阿莫代伊和萨姆·奥尔特曼等科技领袖推动监管，但当前「失控」「流氓」等修辞误导认知。

Unlocking hidden revenue streams with market models

MIT Tech Review AI · 08/20 17:47

Each day, an airline transports tens of thousands of passengers on hundreds of flights. Often these are not straightforward point-to-point routes, with passengers requiring multiple connections. The airline can consider potentially hundreds of variables to price each of these journeys: demand, seaso

中文介绍 文章介绍航空公司通过市场模型解锁隐藏收入流，考虑数百个变量为多段旅程定价，优化收益管理。

Introducing AI Futures

OpenAI News · 08/20 15:00

Introducing AI Futures, a new OpenAI blog exploring how transformative AI could reshape power, governance, the economy, and individual freedom.

中文介绍 OpenAI推出新博客「AI Futures」，探讨变革性AI如何重塑权力、治理、经济和个人自由。

not much happened today

Smol AI News · 08/20 13:44

**OpenAI** and **Anthropic** expanded their agent platforms with new desktop features, collaborative editing, and composable APIs like Skills and Files API. **OpenAI** rolled out memory and workflow features in the EEA, UK, and Switzerland. **AT&T** revealed that 40% of employee AI usage routes to o

中文介绍 OpenAI和Anthropic扩展代理平台，新增桌面功能、协作编辑及可组合API（如Skills和Files）。OpenAI在欧洲经济区、英国和瑞士推出记忆与工作流功能。AT&T透露40%员工使用相关工具。

[AINews] Death of Params: Z.ai CEO Jie Tang on GLM 5.3 and the new Post-training Scaling Law

Latent Space · 08/20 13:17

Every lab CEO is on X now

中文介绍 智谱AI CEO唐杰在Latent Space访谈中讨论GLM 5.3及新的后训练扩展定律，提出「参数死亡」观点。

Stampli cuts launch hours by 68% using ChatGPT Work

OpenAI News · 08/20 08:00

With a fixed deadline and design resources committed elsewhere, Stampli used Codex and ChatGPT Work to compress weeks of launch production into days.

中文介绍 Stampli使用Codex和ChatGPT Work，将发布生产时间从数周压缩至数天，上线时间缩短68%。

Muse Video leaks 📹, Ramp Router launch 🔀, why Stripe bought OpenRouter 💰

TLDR AI · 08/20 08:00

中文介绍 TLDR AI摘要：Muse视频泄露，Ramp Router推出，Stripe收购OpenRouter的原因分析。

Offering Zero Data Retention for frontier models

OpenAI News · 08/20 03:00

OpenAI reaffirms Zero Data Retention for eligible API customers and previews Private Safety Processing for advanced AI safety without compromising data privacy.

中文介绍 OpenAI重申对符合条件的API客户提供零数据保留，并预览「私有安全处理」功能，在不牺牲数据隐私的情况下实现高级AI安全。

[AINews] Memory prices up 500% in 12 months

Latent Space · 08/19 16:44

the Memory crunch continues - Moore’s Law reversed to 2007 levels

中文介绍 AI新闻：内存价格12个月内上涨500%，摩尔定律倒退至2007年水平，内存短缺持续。