2026-09-22

DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression

👍 146

09/17 08:00

The widespread adoption of long-horizon agents has made model workloads increasingly input-heavy. Although prior work has substantially reduced the cost of long-context computation, prefill remains computationally expensive, and large KV caches continue to strain HBM and SSD capacity and data-transf

中文介绍 针对长程 agent 带来的输入密集型负载，在已有长上下文计算优化之上进一步压缩 KV cache，以缓解仍然高昂的 prefill 计算开销以及大 KV 缓存对 HBM、SSD 容量与带宽的压力，把 KV 缓存压缩推向新极限。

Can MiniMax-H3 Reason About the Physical World? An Evaluation of Omni-Modal Generative Model

👍 116

09/16 08:00

Recent Omni-Modal Generative Models (Omni-Models) have advanced content generation toward unified modeling of text, images, video, and audio. MiniMax-H3 exemplifies this transition by combining multimodal context understanding with joint audio-visual generation in a shared latent framework. Its unif

中文介绍 评测全模态生成模型 MiniMax-H3 是否真正具备物理世界推理能力。该模型在共享隐空间中统一文本、图像、视频与音频的理解，并做联合音视频生成。工作据此考察统一多模态生成范式的物理常识与推理边界。

When EOS Tokens Disagree: Understanding Length Inflation in On-Policy Distillation

👍 96

09/17 08:00

We study length inflation in on-policy distillation (OPD), where student responses can become excessively long and even exhaust the generation budget. We identify termination-token mismatch between base students and post-trained teachers as an important source of this behavior. Across Qwen3, Llama,

中文介绍 研究 on-policy distillation（OPD）中的长度膨胀现象：学生回答过长甚至耗尽生成预算。作者发现基座学生与后训练教师之间的终止 token（EOS）不匹配是重要成因，并在 Qwen3、Llama 等模型上验证了该机制。

CodeMidas: Scaling Agentic Coding RL Environments from Code Itself

👍 88

09/18 08:00

Training capable coding agents via reinforcement learning (RL) requires diverse tasks with reliable verifiers. Open-source codebases offer a rich source of such tasks, while existing methods typically rely on development artifacts such as issues and commits, limiting the range of tasks that can be e

中文介绍 用 RL 训练编码 agent 需要多样任务与可靠验证器。CodeMidas 直接从开源代码库本身扩展出可验证的 agentic coding 训练环境，不再依赖 issue、commit 等开发痕迹，从而大幅拓宽可自动构造的任务范围。

Grounded Skill Synthesis from Code at Scale for Agentic Intelligence

👍 86

09/04 08:00

Reusable skills give agents transferable procedural knowledge, making scalable acquisition essential for extending agents beyond prior experience. Existing methods face two limitations: trajectory-based synthesis requires interactions with specific environments, while document-derived skills may lac

中文介绍 面向 agent 可复用技能的规模化获取，提出从代码中大规模合成有据可依（grounded）的技能。现有方法或需与特定环境交互采集轨迹，或由文档推导技能而缺乏可验证性；该方法从代码出发提供可迁移的程序性知识。

EvoOntology: A Self-Evolving Ontology Layer for Data Agents

👍 69

09/14 08:00

Data agents aim to fulfill natural-language instructions over heterogeneous data, including tables, files, and databases. However, data agents face a challenging agent-data gap: heterogeneous data resides outside the agent, while the agent can access it (e.g., column names and file paths) only throu

中文介绍 数据 agent 需在表格、文件、数据库等异构数据上执行自然语言指令，但存在 agent-data gap：数据位于 agent 之外，只能通过列名、文件路径等有限元信息访问。EvoOntology 引入自演化本体层来弥合这一鸿沟。

RecreationWorld: Scalable and Verifiable Environments for Hybrid Computer-Use Agents

👍 60

09/18 08:00

Computer-use agents (CUAs) have advanced along two separate lines: graphical interaction and software development through code and the command line. Real digital work requires both, interleaved rather than stacked end to end. We study hybrid CUAs that autonomously decide when to explore an interface

中文介绍 真实数字工作需要在图形界面操作与代码/命令行之间交替进行，而非端到端堆叠。RecreationWorld 为可自主决定何时探索界面、何时写代码的混合 computer-use agent 提供可扩展且可验证的训练环境。

JEPA-Anything: Learning Predictive Models across Different Worlds

👍 58

09/17 08:00

World modeling enables intelligence to anticipate consequences, guide interventions, and learn from interaction. Yet predictive models remain domain-specific: can a common learning principle support world modeling across radically different systems? We introduce JEPA-Anything, a domain-agnostic fram

中文介绍 世界模型能让智能体预测后果、指导干预并从交互中学习，但现有预测模型多为领域专用。JEPA-Anything 提出领域无关框架，用统一学习原则在差异极大的系统上训练预测模型，探索跨世界建模的可行性。

Verifiable Social Reasoning for LLM Assistants

👍 51

09/15 08:00

LLM assistants are widely used for daily social advice, yet evaluating their social reasoning in such consultation settings remains challenging since (i) it requires setups where the assistant learns about social situations from subjective user narratives, and (ii) social properties, such as others'

中文介绍 LLM 助手常被用于日常社交咨询，但其社交推理难以评估：需从主观的用户叙述中理解社交情境，且涉及他人心理状态等难以验证的社交属性。该工作提出可验证的社交推理评估方案，用于此类咨询场景。

Self-Evolving Search Index

👍 47

09/17 08:00

Information retrieval is increasingly important as LLM agents tackle complex tasks involving diverse information needs. Because retrieval relies on an index that represents each document through index keys, retrieval quality depends heavily on how effectively these keys expose the knowledge containe

中文介绍 检索质量取决于索引键能否有效暴露文档中的知识，而 LLM agent 面对的信息需求日益复杂。该工作提出自演化搜索索引，使索引键随语料与任务动态演进，从而提升面向 agent 的检索效果。

RiskChainBench: A Benchmark for Obfuscated Platform Message Restoration and Evidence-Grounded Web Investigation

👍 42

09/15 08:00

Platform abuse campaigns conceal redirection instructions with emojis, homophones, character decomposition, and redundant symbols, then route users through disguised links to services associated with pornography, fraud, gambling, or illicit transactions. Existing benchmarks evaluate obfuscated text

中文介绍 平台滥用活动用 emoji、谐音、字形拆解和冗余符号隐藏跳转指令，再以伪装链接导流至色情、诈骗、赌博等非法服务。已有基准只测混淆文本识别，RiskChainBench 面向混淆平台信息还原与基于证据的网页调查，覆盖完整链路。

RetireOPD: Self-Retiring On-Policy Distillation for Agentic Reinforcement Learning

👍 41

09/17 08:00

Multi-turn agents trained with reinforcement learning (RL) receive a single scalar reward per trajectory, which motivates self on-policy distillation (OPD) to supply dense token-level supervision from a self-teacher with privileged task skills, letting a skill-free student internalize them. This rec

中文介绍 多轮 agent 的 RL 训练每条轨迹只有一个标量奖励，self on-policy distillation 借助具备特权技能的自教师提供 token 级稠密监督。RetireOPD 让蒸馏过程自我退场，避免学生长期依赖教师，兼顾稠密监督与 agentic RL 效果。

Reflect, Revise, Reuse: Training-Free Skill Evolution for GUI Agents

👍 38

09/15 08:00

GUI agents execute long-horizon tasks on dynamic graphical user interfaces, where pop-ups, delayed loads, and relocated widgets routinely invalidate plans fixed before execution. Recent agent-skill frameworks encapsulate reusable procedural knowledge to mitigate this, yet existing skill designs are

中文介绍 GUI agent 在动态界面上执行长程任务时，弹窗、延迟加载与控件移位常使预设计划失效。该工作提出「反思—修正—复用」的免训练技能演化框架，让已有可复用技能在执行中自动更新，无需额外训练。

WeVisDoc: From Coverage to Capability for Robust End-to-End Document Parsing

👍 35

09/17 08:00

Document parsing converts document images into structured content and requires reliable performance across diverse layouts and acquisition conditions. Yet training corpora are biased toward common document types and clean digital pages, while expanding coverage alone does not specify how to address

中文介绍 文档解析要把文档图像转为结构化内容，并需在多样版式与采集条件下保持可靠，而训练语料偏向常见文档类型与干净数字页，单纯扩大覆盖面不足以解决问题。WeVisDoc 从覆盖转向能力，构建更鲁棒的端到端解析方案。

When2Think: Learning Difficulty-Aware Length Control for Efficient Hybrid Reasoning Models

👍 33

09/17 08:00

Large Reasoning Models (LRMs) achieve strong performance on complex tasks but exhibit systematic inefficiency: they often overthink easy problems and underthink hard ones. Existing approaches based on uniform length penalties or rigid routing incur an efficiency tax, trading reduced computation on e

中文介绍 大型推理模型常对简单问题过度思考、对困难问题思考不足。When2Think 提出难度感知的长度控制方法，让混合推理模型按题目难度自适应决定思考长度，避免统一长度惩罚或刚性路由带来的效率损失。

OmniVChat: Synthesizing, Benchmarking, and Training for Native Audio-Visual Dialogue

👍 30

09/18 08:00

We define OmniVChat (Omni Video Chat) as the task of native audio-visual dialogue between a user and an omni model. In OmniVChat, omni models directly and simultaneously receive audio and video from a user and return text. The user's query is embedded in the audio and video, without a separate text

中文介绍 定义 OmniVChat（原生音视频对话）任务：omni 模型同时接收用户的音频与视频并直接返回文本，查询内嵌于音视频之中，无需单独文本输入。工作覆盖数据合成、基准构建与模型训练三部分。

What Does Privileged Information Add to On-Policy Self-Distillation?

👍 30

09/17 08:00

On-policy self-distillation (OPSD) lets a language model learn from a frozen copy of itself that sees an answer or a worked solution. Giving the teacher this extra information seems to offer the student more to learn, but how much does it add beyond distillation itself? To isolate that contribution,

中文介绍 on-policy self-distillation 让学生向一个能看到答案或解题过程的自身冻结副本学习，但特权信息究竟在蒸馏之外贡献多少并不清楚。该工作通过隔离实验量化这部分增益，厘清其真实作用。

UFO: Chain-of-Evaluation for Omni-Condition Alignment in Multi-Modal Image Generation

👍 30

09/17 08:00

Multi-modal image generation, particularly subject-driven customization, has garnered growing attention in recent years. Despite the rapid advancement of generative models, their evaluation remains largely lagging. Existing methods, whether embedding-based or Multi-modal Large Language Model (MLLM)-

中文介绍 多模态图像生成尤其是主体驱动定制发展迅速，但评价方法明显滞后，基于嵌入或 MLLM 的现有方案难以刻画多样条件的对齐程度。UFO 提出链式评估（Chain-of-Evaluation）框架，用于多模态条件对齐的细粒度评测。

Paint-Anything: Unified Any-Color Control for Image Generation and Editing

👍 29

09/17 08:00

Professional design requires any-color control: the ability to specify an object's target color with any 24-bit hex value for image generation and editing. Prior work has explored color generation, editing, and colorization, but often relies on dedicated color representations or specialized inferenc

中文介绍 专业设计要求任意颜色控制，即用任意 24 位 hex 值指定物体目标颜色。Paint-Anything 提出统一的任意颜色控制方法，同时支持图像生成与编辑，无需依赖专用颜色表示或专门设计的推理模块。

Sample Count Is Not Enough: Candidate-Generation Strategy Shapes the Energy and Performance of LLM Test-Time Scaling

👍 29

09/16 08:00

Test-time scaling can improve large language model reasoning by generating and combining multiple candidate responses. In sampling-based methods, the inference budget is often described by the number of generated candidates, N. However, N tells us how many candidates are generated, not how they are

中文介绍 测试时扩展常以候选数量 N 描述推理预算，但 N 只说明生成多少候选，并不反映候选如何生成。该研究指出候选生成策略显著影响测试时扩展的能耗与推理性能，仅用采样数衡量预算并不充分。

You're Already a Meat Proxy

@obie · 13.1K 粉丝 · 157.2K 阅 · 738 赞 · 98 转

09/21 05:02

I’m excited about a future that I suspect will be very hard on many long-time friends and certain people that I love. The same technology that is allowing me to be more successful than ever is

中文介绍 博主对 AI 时代的一则反思：同一套让他效率远超以往的技术，也可能让许多老友和所爱之人陷入困境。属于观点类长文，讨论技术进步背后的个人与关系代价，没有具体工具或方法论。

I reverse-engineered Instinct's memory. Here's exactly how it works

@DhravyaShah · 62.1K 粉丝 · 120.2K 阅 · 509 赞 · 16 转

09/21 02:49

Instinct has taken the world by storm over the last two weeks - it's one of the best imessage assistants I've used. as with every product, I decided to reverse-engineer Instinct's memory to find out

中文介绍 博主逆向拆解 iMessage 助手 Instinct 的记忆机制，分析它两周内爆红的原因，以及记忆究竟如何存取与调用。属于产品逆向工程类技术分析，适合想给自己的 AI 助手搭记忆层的人参考。

Codex vs Claude for app development

@twostraws · 120.7K 粉丝 · 77.1K 阅 · 504 赞 · 42 转

09/19 22:30

I use both Codex and Claude for app development work, both with the highest-tier subscription. Here's how I've found the two compare, plus some tips for success… TL;DR I use both agents extensively,

中文介绍 博主同时用 Codex 和 Claude 做 App 开发，且都开了最高档订阅，对比两者在实际开发中的表现差异，并附上提高成功率的使用技巧。属于长期实战后的经验对比，而非单次跑分。

A Beginner's Guide to Jev

@omarsar0 · 320.6K 粉丝 · 73.5K 阅 · 523 赞 · 60 转

09/21 04:44

As you may have seen recently, Jev (a generalist System One model) took Twitter by storm. Many misleading examples are everywhere, but they don't really show Jev's real capabilities. Hence, we

中文介绍 Jev（通用 System One 模型）近期刷屏，网上流传的示例多有误导。博主基于其真实能力整理了一份入门指南，讲清 Jev 到底是什么、能做什么，适合被各种演示搞晕的人先看这篇。

Jev, System One models, and the future of computer use

@trycua · 19.0K 粉丝 · 59.5K 阅 · 501 赞 · 39 转

09/20 06:27

What we've been building with decision models at Cua - by @ddupont808 and @francedot Over the last few days we've deprioritized some other shiny objects because we've all been Jev-crazy. TypeSafe

中文介绍 Cua 团队分享他们做决策模型（decision models）的思路：过去几天为 Jev 让路、搁置其他项目，并讨论 System One 模型与 computer use 的未来走向。偏产品方向与路线判断，不是教程。

How to Use Jev to Build a 24/7 HFT Trading System

@RohOnChain · 72.5K 粉丝 · 57.7K 阅 · 518 赞 · 54 转

09/19 22:05

I will break down the exact framework to build a millisecond speed 24/7 HFT trading system with Jev, along with the exact resources that helped me personally. Let's get straight to it. Bookmark This -

中文介绍 一份用 Jev 搭建 7×24 小时毫秒级 HFT 交易系统的框架拆解，包含具体实现思路和作者亲测有用的资源清单。属于把新模型直接落到高频交易场景的实操教程，偏进阶向。

Jev-as-a-Judge for Agent Evals

@LangChain · 265.9K 粉丝 · 46.6K 阅 · 504 赞 · 68 转

09/20 07:31

By Daniel Shea and Seán Roche Key takeways: Jev is a fundamentally different kind of evaluator. It returns typed answers directly instead of generating text like an LLM judge. Jev was dramatically

中文介绍 LangChain 团队测试用 Jev 充当 agent 评测的裁判：它不像 LLM judge 那样生成文本，而是直接返回带类型的结构化答案，评测表现明显更好。属于 agent evals 的方法论更新。

Jev 101: what it is, and what you can build with it

@manish_fp · 846 粉丝 · 44.1K 阅 · 512 赞 · 59 转

09/19 22:59

I have read 10+ articles, watched 5 hours of content to understand what actually Jev is and then chose to write this. It is not a chatbot, It’s a whole new kind of AI, more like a classifier which

中文介绍 博主读了 10 多篇文章、看了 5 小时视频，才搞清 Jev 到底是什么：它不是聊天机器人，更像一个分类器式的新 AI。文章用直白方式解释 Jev 及可搭建的东西，适合零基础入门。

Build your own Jev (100% local)

@_avichawla · 77.5K 粉丝 · 37.8K 阅 · 593 赞 · 80 转

09/20 14:46

Everything you need to turn an open-source LLM into a fast, local decision engine without retraining it. It covers next-token scoring, fixed choices with probability distributions, SGLang, and a

中文介绍 一篇不重新训练模型、把开源 LLM 改造成本地高速决策引擎的教程，覆盖 next-token 打分、固定选项下的概率分布、SGLang 等要点，可 100% 本地跑通，适合想自建 Jev 式系统的人。

A Swarm of Fruit Flies, Building a Civilization On-Chain

@murmur_arc · 983 粉丝 · 11.8K 阅 · 553 赞 · 137 转

09/20 21:17

Prologue — No LLMs, Only Neurons This project is called murmur, and it is alive right now on Arc mainnet (Chain ID 5042), settling in real USDC. To date the flies have completed more than 23,000

中文介绍 项目 murmur 用一群果蝇神经元代理在链上自主演化，完全不用 LLM，目前运行在 Arc 主网（Chain ID 5042），以真实 USDC 结算，已完成逾 2.3 万次交互。属于链上多智能体实验，偏概念与叙事。

Jev: System One models for Prod, not God — with Diogo Almeida, CEO, TypeSafe AI

Latent Space · 09/22 06:13

The definitive Jev podcast with its lead creator.

中文介绍 Latent Space 发布与 Jev 主创者的播客节目，嘉宾为 TypeSafe AI 首席执行官迪奥戈·阿尔梅达，主题涉及面向生产环境而非追求通用智能的「系统一」模型。

Pruning LLMs Like a Physicist: Block Removal as an Ising Optimization Problem

Hugging Face Blog · 09/21 21:44

中文介绍 Hugging Face 博客发文，提出以物理学视角修剪大语言模型，将模块移除问题建模为伊辛优化问题来求解。

Advisory Group on Mathematics and Artificial Intelligence

OpenAI News · 09/21 20:00

OpenAI is working with an independent Advisory Group on Mathematics and Artificial Intelligence to guide the review and communication of emerging AI results.

中文介绍 OpenAI 宣布与独立的「数学与人工智能咨询小组」合作，为新兴 AI 成果的评审和对外沟通提供指导。

Higgsfield AI ships new video features in a day with GPT-6 Astra

OpenAI News · 09/21 20:00

With GPT-6 Astra, Higgsfield AI makes video ad creation easier for small businesses and brings new creative tools to market faster.

中文介绍 OpenAI 表示，Higgsfield AI 借助 GPT-6 Astra 在一天内上线新的视频功能，让中小企业更易制作视频广告，并加快创意工具推向市场。

How we made the first comprehensive map of deaths along the US border’s “virtual wall”

MIT Tech Review AI · 09/21 20:00

Our 15-month investigation into death and surveillance along the US-Mexico border began with a simple question: Why did so many people die near government surveillance towers meant to help track and apprehend them? This story is part of Dying on Camera, a collaboration between MIT Technology Review

中文介绍 MIT 科技评论披露历时 15 个月的调查，首次绘制美墨边境「虚拟墙」沿线的死亡地图，起因是为何大量人员死在用于追踪和抓捕的政府监控塔附近。

4 ways to address the failures we found along the US border’s “virtual wall”

MIT Tech Review AI · 09/21 20:00

MIT Technology Review today published our investigation into how many people have died near the “virtual wall” of surveillance towers that the US government has installed along the US-Mexico border. We found cases of people who walked undetected through areas surveilled by advanced, AI-enabled tower

中文介绍 MIT 科技评论在发布美墨边境监控塔「虚拟墙」调查后，提出四项政策建议以应对所发现的问题；调查发现有人未被察觉地穿过先进监控区域。

The US spent billions on border surveillance. Why can’t it catch people before they die?

MIT Tech Review AI · 09/21 20:00

When José Morales Bernal crossed the border into the United States on April 8, 2024, the day before his 32nd birthday, it should have triggered a chain of technological alerts and human responses. As he walked through the desert in southern New Mexico that morning, he was within range of three surve

中文介绍 MIT 科技评论调查指出，美国在边境监控上花费数十亿美元，却未能赶在人员死亡前发现他们。2024 年 4 月 8 日，何塞·莫拉莱斯·贝尔纳尔在 32 岁生日前一天越境新墨西哥州沙漠，本应触发技术警报与人工响应。

She died at the San Diego border. A surveillance camera was in plain sight

MIT Tech Review AI · 09/21 20:00

She had only walked for a couple of hours, and already she was lost. It was early afternoon on Sept. 14, 2025, when 30-year-old Graciela Gómez Hernández crossed the border from the eastern edge of Tijuana into Southern California, sending voice messages to her mother and sister as she walked. This s

中文介绍 2025 年 9 月 14 日，30 岁的格拉谢拉·戈麦斯·埃尔南德斯从蒂华纳东部越境进入南加州，仅走了几个小时便迷路，最终在圣地亚哥边境身亡，而监控摄像头就在近旁。

Building standards for the next phase of AI

OpenAI News · 09/21 18:00

OpenAI outlines a path to shared global AI standards, calling for coordinated evaluation, reporting, and governance to improve safety.

中文介绍 OpenAI 阐述推进全球统一 AI 标准的路径，呼吁在评估、报告与治理方面加强协调，以提升安全性。

Expanding OpenAI Academy with new learning paths

OpenAI News · 09/21 15:00

Explore new OpenAI Academy learning paths for employees, developers, leaders, educators, and students to build and demonstrate practical AI skills.

中文介绍 OpenAI 宣布扩展 OpenAI Academy，新增面向员工、开发者、管理者、教育者和学生的学习路径，帮助培养并展示实用 AI 技能。

How V7 gives AI agents institutional memory

OpenAI News · 09/21 08:00

Using GPT-5.6, V7 turns scattered company files into context agents can use to complete complex, source-linked work.

中文介绍 OpenAI 介绍 V7 如何运用 GPT-5.6，将企业内零散文件转化为智能体可用的上下文，从而完成复杂且可溯源的工作。

tokenizers v1: encode, decode and scaling, measured

Hugging Face Blog · 09/21 08:00

中文介绍 Hugging Face 博客发布 tokenizers v1 版本，介绍其在编码、解码与扩展性方面的实测表现。

Muse connectors 🤖, Meta SAM 3.1 🖼️, Gemini hacks companies 🔓

TLDR AI · 09/21 08:00

中文介绍 TLDR AI 汇总当日 AI 动态，涉及 Muse 连接器、Meta 的 SAM 3.1 图像模型，以及 Gemini 被用于入侵企业相关消息。

[AINews] Here are 6 Clones of Jev in 2 days

Latent Space · 09/19 13:48

Imitation is the sincerest form of Flattery

中文介绍 Latent Space 的 AINews 报道称，两天内出现了 6 个 Jev 的克隆版本，并称模仿是最真诚的赞美。

Introducing the Australian Youth Safety Blueprint

OpenAI News · 09/18 20:00

OpenAI introduces the Australian Youth Safety Blueprint, a six-pillar roadmap for safer AI experiences that protect and empower young people.

中文介绍 OpenAI 发布「澳大利亚青少年安全蓝图」，提出包含六大支柱的路线图，旨在打造更安全的 AI 体验，保护并赋能年轻人。