2026-09-30

Scaling Properties of Same-Family On-Policy Distillation

👍 197

09/25 20:00

*Reinforcement learning (RL)* can induce substantial reasoning capabilities in large language models (LLMs), but how much of this capability transfers across model scales, and how quickly, remains unclear. We study the scaling properties of *on-policy distillation (OPD)* across *weak-to-strong*, *sa

中文介绍 针对强化学习（RL）在大型语言模型（LLMs）中引入推理能力的能力转移速度进行研究，论文分析了在弱到强的模型规模中，在线策略蒸馏（OPD）的扩展特性。实验结果表明，OPD能有效迁移推理能力，提高跨模型间的协同效果。意义：为RL与LLMs结合的进一步发展提供理论支持，促进智能体及推理系统的进步。

Omni-IO Skills: Harnessing Your Agent Omni-Native

👍 156

09/24 20:00

General-purpose agents can plan, reason, and act over long horizons, yet their production capabilities remain fragmented across text, images, audio, video, documents, 3D assets, and code. Extending a foundation model to additional modalities ties capability growth to costly model updates, while asse

中文介绍 论文探讨了通用智能体在多个模态中（文本、图像、音频等）规划、推理和行动的能力。提出了一种扩展基础模型的策略，旨在简化更新过程，支持多模态间的灵活互动。通过实验，模型在多个模态上的协同能力显著提升，为多模态智能体的应用打下基础。意义：推动了跨模态智能体能力的整合，应用于更复杂的任务。

VoxMem: Benchmarking Multimodal Memory in Large Audio Language Models

👍 125

09/25 20:00

Spoken conversational systems must recover information from prior interactions (i.e., memory), yet relevant information in speech extends beyond what was said to who said it, how it was spoken, and what was audible, information that exists only in the audio signal and cannot be recovered from a tran

中文介绍 针对口语对话系统中的记忆问题，论文提出了一种新的多模态记忆基准框架VoxMem，强调了从语音中提取上下文信息的重要性。通过对模型的评估，展示了其在捕捉说话者、语气和可听信息方面的效益，支持更复杂的对话管理。意义：为口语对话系统的记忆增强提供了研究基础，推动智能体在对话场景中的应用。

Periodic Weak Spots: Phase Sensitivity from Chunked KV-Cache Compression

👍 92

09/27 20:00

Chunked KV-cache compression reduces the memory and attention costs of long-context inference by compressing windows of consecutive tokens into fewer cache entries at a fixed stride. Such compression also introduces a new positional coordinate: a token's phase, or its position relative to compressio

中文介绍 提出了一种基于分块KV-cache压缩的方法，旨在通过在固定步长内将连续标记窗口压缩成更少的缓存项来降低长上下文推理的内存和注意力成本。该方法引入了新的相对位置标志：标记的相位，有效提升了模型在推理过程中的处理效率。意义：在提高长上下文推理效率的同时，推动了模型在复杂推理任务中的应用。

Follow the Entities: A Corpus Map for Agentic Search

👍 68

09/28 20:00

Answering questions and completing tasks over large document collections often requires connecting evidence spread across multiple documents, such as a project's approval recorded in one, its requirements in another, and its latest status in a third. Recent LLM agents approach this by iteratively se

中文介绍 探索了如何在大规模文档集合中回答问题和完成任务，提出了一种基于证据连接的搜索策略。通过对LLM智能体的迭代设计，模型能有效整合分散在多个文档中的信息，改善任务完成的准确性和效率。意义：为文档检索与智能体的有效结合提供了新方案，推动信息检索技术的发展。

Beyond Dyadic Memory: Interaction-Aware Multimodal Memory with Adaptive Agentic Retrieval for Multi-Party Spoken Conversations

👍 62

09/25 20:00

Long-term memory enables agents to accumulate information and reason across sessions, yet existing research primarily focuses on dyadic text or image-text conversations, leaving long-term memory for multi-party spoken conversations underexplored. This setting requires preserving conversational conte

中文介绍 针对多方口语对话中的长期记忆研究，提出了一种交互感知的多模态记忆模型，增强了信息积累与推理能力。通过自适应智能体检索，模型能够更有效地保持对话上下文和信息，显著提升多方交互场景下的反应能力。意义：推动了在复杂对话环境下记忆机制的研究，为智能助手的发展提供支持。

LLMs are General Asynchronous Agents

👍 61

09/27 20:00

Modern LLMs are increasingly capable as autonomous agents, but they follow sequential interaction cycles: read, think, reply or call tools, repeat. Many real-world use cases are not sequential: voice assistants, embodied agents, and monitoring systems receive new inputs while they think or perform a

中文介绍 现代LLMs作为自主智能体正在快速发展，论文指出其在执行序列交互循环时的局限性，提出了一种非线性处理的方法，使智能体能够在处理信息的同时接收新输入。该方法有效提高了智能体在动态环境中的适应性。意义：为智能体在实时响应和多任务处理中提供了新思路，影响智能系统的应用实践。

LongCat-DeepResearch Technical Report

👍 54

09/27 20:00

We present LongCat-DeepResearch, a deep research system that combines an enhanced LongCat model with a multi-agent workflow for producing comprehensive, evidence-grounded reports. The workflow separates global planning from detailed investigation and coordinates revision at the section level. Multip

中文介绍 构建了LongCat-DeepResearch，一个结合增强型LongCat模型与多智能体工作流程的深度研究系统。该系统能够将全局规划与细致调查相分离，并在各个部分进行协作修订，从而高效生成基于证据的综合报告。意义：为通过多智能体方法提升研究效率奠定基础，影响信息整合和处理的领域。

EngiWorld: What Can Frontier Agents Deliver in Professional Engineering Environments?

👍 49

09/28 20:00

Autonomous agents have made rapid progress in general-purpose computer use, but reliable automation of professional industrial engineering remains out of reach, as engineering workflows demand reasoning over geometric and physical constraints and dependencies preserved across software and design sta

中文介绍 论文探讨了自主智能体在专业工程环境中的应用，指出尽管进展显著，但在复杂几何和物理约束的可靠自动化工程工作流依然面临挑战。通过分析现有系统的缺陷，为改进智能体的自动化能力提供指导。意义：推动了智能体在工业与工程领域的应用研究，提升工程效率。

APM-Bench: Benchmarking Cross-session Persistent Memory for Egocentric Streaming Video Assistants

👍 29

09/28 20:00

To serve as real-world personal assistants, streaming video models need persistent memory that retains past experiences for later use. Yet existing streaming benchmarks and methods often focus on individual continuous videos or short clips, overlooking that real-world interactions are often intermit

中文介绍 为了作为现实世界的个人助理，流媒体视频模型需要具备持久记忆，以便保留过去的经验供后续使用。论文提出了APM-Bench基准，强调持续视频与短视频的差异，从而更好地服务于真正的人机交互场景。意义：推动流媒体智能助理的研究与应用，拓展人工智能在视频处理领域的发展。

SoL-Refiner: Speed-of-Light One-Step Refinement for High-Resolution Video

👍 28

09/28 20:00

High-resolution video generation is expensive, as its cost grows rapidly with the number of spatiotemporal tokens. A practical alternative first generates a lower-resolution video and then applies a refiner, but conventional multi-step refinement introduces a second sampling bottleneck. We present S

中文介绍 针对高分辨率视频生成过程中的成本问题，论文提出SoL-Refiner方法，通过第一步生成低分辨率视频并应用精细化技术，降低生成过程中的抽样瓶颈，提升了生成效率。意义：为视觉生成技术的发展提供新方法，推动高效的视频生成应用。

Asking for What Was Never Requested: Horizontal and Vertical Proactivity in Agents

👍 27

09/28 20:00

An agent that uses tools typically responds to what the user explicitly asks, yet completing the task may require information the user never requested. Work on proactive agents mainly studies whether and when an agent should act on its own, not what information it should pursue. We study a distinct

中文介绍 研究了一种递归推理模型，在循环Transformer结构中调度推理过程。提出的模型能够根据更新的需要动态调整推理方式，提高推理的灵活性和效率，克服了传统模型的局限性。意义：为推理机制提供新的视角，有助于推动智能体在复杂推理任务中的表现。

Context Language Models

👍 23

09/28 20:00

We introduce Context Language Models (CLMs), language models that natively manage their own context. We implement this by treating the context as a file and allowing the model to make unrestricted updates to this file. This allows the model to learn what is most important to maintain in context, and

FocusVTC: Efficient and High-Performance Visual Text Compression with Adaptive Resolution

👍 20

09/28 20:00

Long-context reasoning in large language models incurs substantial computation and memory costs. Visual text compression (VTC) reduces input length by rendering text as images, but fixed-resolution rendering creates a compression-performance trade-off: low DPI saves tokens at the expense of legibili

StoryEngine: A State-Grounded Agentic Framework for Video Storytelling

👍 16

09/26 20:00

Despite recent progress in agentic multi-shot video generation, producing coherent and consistent long-form stories remains challenging. Existing agentic pipelines typically rely on textual shot plans or previously generated pixels, yet lack an explicit mechanism for propagating the consequences of

Omni-Decision: Evidence-Ledger Planning for Omni-Modal Agents

👍 16

09/23 20:00

Omni-modal agents must seek evidence across video, audio, web pages, and computation to answer questions. Their main bottleneck is planning: noisy multimodal observations accumulate in conversation history and disrupt later decisions, while multimodal models have limited capacity for multi-step plan

Reasoning with Image Generation

👍 14

09/14 18:29

Chain-of-thought reasoning has revolutionized natural language processing by enabling large language models (LLMs) to decompose problems into intermediate steps before answering. Yet confining reasoning to the textual domain presents limitations for tasks requiring direct manipulation of visual repr

AutoRef: Harness Optimization for Agentic Multi-Reference Image Generation

👍 13

09/27 20:00

Recent image generation models can take multiple reference images as input and combine them into a new image. However, multi-reference image generation remains challenging: models may omit or duplicate subjects from the references, or produce images in which multiple subjects appear unnaturally past

Chinese-Jev: Bringing System One Model to Chinese-Language Tasks

👍 12

09/28 20:00

System One models such as Jev offer an efficient alternative to generative language models for tasks that require decisions rather than open-ended responses. However, existing Jev models exhibit limited Chinese-language decision accuracy, restricting their utility in both general and specialized set

Scheduling Recursive Reasoning in Looped Transformers

👍 11

09/28 20:00

Recurrent reasoning models have attracted growing attention for scaling test-time computation, typically by iteratively refining latent states with shared parameters. However, these models apply each learned update with a fixed unit scale, which can be conservative when updates make persistent progr

Teamily vs. OpenAI Dots: The Vision "Team of Dots" is Already Here, Using by Million of Users

@Teamily_AI · 1.6K 粉丝 · 749.2K 阅 · 512 赞 · 170 转

09/29 16:00

OpenAI just shared its vision for Dots: an always-on personal agent today, evolving over time into teams of specialized agents working together. We think that is exactly where AI is going. It is also

中文介绍 博主探讨了 OpenAI 发布的 Dots 计划，该计划旨在实现个性化代理的不断发展，最终形成团队协作的专门代理。此内容反映出“团队代理”在 AI 发展中的趋势，显示出用户需求正在向高度协作的智能助手转变。

Meet your Team Bots

@n2parko · 5.7K 粉丝 · 274.2K 阅 · 533 赞 · 38 转

09/28 16:03

Today we're introducing Team Bots, and it's likely the biggest change to Grok Bot since launch. I wanted to share a bit about why we built it and how these new digital colleages might change the way

中文介绍 博主推出了 Team Bots，这是 Grok Bot 自发布以来最大的变革，介绍了这些新型数字同事的构建初衷及其可能改变的工作方式，反映出智能协作工具在职场中的发展潜力。

Free the models: harness design at the frontier

@pirroh · 16.2K 粉丝 · 234.6K 阅 · 507 赞 · 78 转

09/29 12:48

Model routers are everywhere right now, but they have a fundamental limitation. No matter if based on advanced heuristics or a small model that reads each turn and picks which LLM to use, a router

中文介绍 博主讨论了模型路由器的局限性以及面临的挑战，强调了设计在 AI 发展中的重要性。这种观点表明，智能模型的选择和运用仍需深入探索，反映出当前技术进步中的制约因素。

History Rhymes: Large Language Models Off to a Bad Start?

@michaeljburry · 2.1M 粉丝 · 120.2K 阅 · 519 赞 · 84 转

09/30 11:04

This 145-year-old case study – presented at the Smithsonian Institute no less - provides a potentially devastating critique of today’s Large Language Models and the spending behind them. Michael Burry

中文介绍 博主引用了一份历史案例研究，批判当前大语言模型的表现和背后的投资。这一讨论引发了对大语言模型有效性的新反思，触及了业界对其真实价值的质疑。

The FDE boom is a product gap

@EugenAlpeza · 2.5K 粉丝 · 96.2K 阅 · 539 赞 · 50 转

09/29 10:07

The industry is hiring its way around a missing product. Every enterprise trying to put agents to work runs into the same problem: agents don’t know how the company operates, and the operators who do,

中文介绍 博主指出行业在推动智能代理应用过程中面临的产品缺口，强调代理缺乏对公司运作的理解。此观点反映出企业在部署 AI 代理时的普遍挑战，对行业发展形成警示。

CC, our experimental AI agent built for families

@GoogleAI · 2.5M 粉丝 · 69.4K 阅 · 508 赞 · 36 转

09/28 14:54

Managing a household means juggling busy schedules, crowded inboxes, and endless to-do lists. It requires context switching, multitasking, and the most precious resource of all: time. Earlier this

中文介绍 博主介绍了针对家庭管理的实验性 AI 代理 CC，该代理旨在帮助处理繁忙的生活和事务安排。此内容展现了 AI 在家庭生活中的实用价值，预示着智能助手市场的潜在扩大。

i made 10+ motion videos with opus 5.5 in 3 days. here's the whole pipeline (open source)

@RaphaelAubryy · 393 粉丝 · 40.9K 阅 · 507 赞 · 42 转

09/28 05:25

Most people trying motion design with Opus 5.5 hit the same wall: ↳ the first render is "mid": centered text, a gradient, everything fading in ↳ the music and the picture live separate lives ↳ one

中文介绍 博主分享了使用 Opus 5.5 制作运动视频的全流程，强调了设计在创作过程中的常见障碍，并提供了开源解决方案。这透露出对高效设计工具和流程的持续探索需求。

How I Make $10K Launch Videos for $0 with Opus 5.5 (Full Course)

@twoclipping · 7.5K 粉丝 · 40.5K 阅 · 510 赞 · 35 转

09/28 01:42

im open sourcing my entire workflow for making the motion designs that studios charge $5,000 to $15,000 for this uses 0 mcp, 0 external tools and im not selling you any subscription you dont need

中文介绍 博主开源了制作高质量启动视频的完整工作流程，指出这项工作无需高昂的外部工具和订阅。反映出业界对高效、低成本创作工具的追求，展示了 DIY 文化的兴起。

DeepSeek Harness for Desktop Is Here

@DeepSeekHarness · 989 粉丝 · 35.4K 阅 · 1.0K 赞 · 69 转

09/30 12:13

Today, we’re releasing the DeepSeek Harness v0.2 preview, with a desktop app available for macOS and Windows. You can install it here and follow the instructions to get started. This release comes

中文介绍 博主宣布 DeepSeek Harness v0.2 预览版发布，包含适用于 macOS 和 Windows 的桌面应用程序，展示了在桌面环境中推动智能工具可用性的努力，表明这一领域的快速进展。

Building a design system for coding agents

@arlanoska · 3.5K 粉丝 · 25.4K 阅 · 503 赞 · 29 转

09/29 14:26

You ask an agent to build a page. Then you fix the spacing, swap the inputs and explain which button it should use. The next agent makes different choices, so you do it all again. If your repo has

中文介绍 博主探讨了为编码代理构建设计系统的过程，强调了代理之间的协作和设计的一致性需要。这反映出在复杂项目中，智能代理的协同工作仍需改进。

Why Dwarkesh is Wrong about Computer Use + How OpenAI shipped its Jev competitor in 1 Week

Latent Space · 09/30 18:23

Our DevDay coverage - the first pod on the DevDay lineup - dives in with the leaders of OpenAI’s CUA team and API platform.

中文介绍 在OpenAI的CUA团队和API平台领导者的访谈中，讨论了DevDay活动的重要内容及OpenAI的快速创新。

Gemini 4 Argon: our next era of frontier intelligence

DeepMind Blog · 09/30 16:01

中文介绍 DeepMind推出Gemini 4 Argon，标志着前沿智能的下一个时代。

Introducing SynthID Bio

DeepMind Blog · 09/30 11:03

Proof of concept for watermarking AI-generated proteins while preserving biological function.

中文介绍 DeepMind展示了SynthID Bio的概念验证，旨在为AI生成的蛋白质添加水印，同时保留生物功能。

“We’re not going to shoot ourselves in the foot” over hack fallout, says OpenAI’s chief research officer

MIT Tech Review AI · 09/30 06:40

Two months after the bombshell news that a swarm of its agents had broken their containment and hacked into the computers of the AI company Hugging Face, OpenAI is still putting out fires. A steady drip of disclosures about other hacks in the weeks since has kept OpenAI in the spotlight and raised s

中文介绍 OpenAI首席研究官表示，在Hugging Face被黑客事件后，OpenAI正在持续应对相关的泄露问题。

Disrupting a coordinated model-distillation campaign

OpenAI News · 09/30 06:30

Learn how OpenAI disrupted a campaign to extract protected model reasoning and is strengthening defenses against adversarial distillation.

中文介绍 OpenAI成功打击了一项提取保护模型推理的统一模型蒸馏活动，并加强了防御措施。

Helping small businesses put AI to work

OpenAI News · 09/30 06:00

OpenAI is partnering with America’s SBDC to expand hands-on AI training and local support for small businesses, alongside a new report on how small teams are using AI.

中文介绍 OpenAI与美国小商业发展中心合作，为小企业提供AI培训与当地支持，并发布相关报告。

[AINews] OpenAI DevDay 2026: Dots, 6.1 Sol, Ultrafast, Decisions API, Agents API, Spaces, Marketplace, and 1.2 Billion ChatGPT WAU

Latent Space · 09/30 01:53

the most confident DevDay yet.

中文介绍 2026年OpenAI DevDay展示了多个新功能，体现出前所未有的信心。

Open TTS Leaderboard: Scalable Evaluation for Multilingual Text-to-Speech and Voice Cloning

Hugging Face Blog · 09/29 20:00

中文介绍 Hugging Face推出了Open TTS排行榜，以便于多语言文本转语音及声音克隆的可扩展评估。

OpenAI Dots 🟡, GPT-6.1 Sol ⚡, software factories 🏭

TLDR AI · 09/29 20:00

中文介绍 OpenAI发布了新的功能包括Dots和GPT-6.1 Sol，提升了软件工厂效能。

NVIDIA Kumo Tabular Sets a New Accuracy-Efficiency Frontier for Tabular Prediction

Hugging Face Blog · 09/29 11:30

中文介绍 NVIDIA Kumo Tabular在表格预测的准确性和效率方面设立了新的标准。

Getting the Source Right, Not Just the Fact: Source-Aware Verification for MCP Agents

Hugging Face Blog · 09/29 09:07

中文介绍 Hugging Face推出了源感知验证方法，以增强MCP代理的准确性。

Making AI an asset, not an expense

MIT Tech Review AI · 09/29 06:43

When customers talk about AI costs, the conversation usually starts with token prices and ends with access to the latest, most capable model in the cloud. Do they always need that level of capability? Not necessarily. But that is often where the conversation goes. As AI moves from experimentation to

中文介绍 当客户讨论AI成本时，通常从代币价格开始，实际上并不一定需要访问最强大的云模型。

Introducing GPT-6.1 Sol

OpenAI News · 09/29 06:00

Meet GPT-6.1 Sol: near-Astra intelligence for coding, computer use, and professional work at one-fifth of Astra’s standard API input and output token prices.

中文介绍 OpenAI推出GPT-6.1 Sol，提供接近Astra智能的能力，成本为Astra标准API输入和输出代币价格的五分之一。

DevDay 2026 Recap

OpenAI News · 09/29 06:00

Explore more than 20 announcements from OpenAI DevDay 2026, including GPT-6 Astra, ChatGPT, Codex, APIs, security, and new tools for builders.

中文介绍 OpenAI DevDay 2026回顾介绍了20多个新发布的项目，包括GPT-6 Astra和新的工具。

[AINews] AMD buys World Labs for $8.2B, as Atlas solves sparse reconstruction problem for robotics, design and more

Latent Space · 09/28 22:55

Congrats team!

中文介绍 AMD以82亿美元收购World Labs，同时Atlas解决了机器人和设计领域的稀疏重建问题。