2026-07-25

AREX: Towards a Recursively Self-Improving Agent for Deep Research

👍 116

07/23 08:00

Deep research requires agents to find answers that jointly satisfy multiple constraints. Discovering such answers is costly, whereas verifying a candidate can often be decomposed into tractable constraint-wise checks. This discovery--verification asymmetry suggests that a research agent should do mo

中文介绍 提出AREX，一种递归自改进的深度研究智能体。利用发现-验证不对称性，通过分解约束进行高效验证，实现递归自我提升，降低搜索成本。

SLAI T-Rex: Full-Parameter Post-training of the DeepSeek-V4 Family on Ascend SuperPOD

👍 56

07/22 08:00

Full-parameter post-training of trillion-parameter-scale MoE models introduces substantial system-level challenges for large-scale distributed training, including severe memory pressure, non-overlapped communication overhead, and inefficient kernel execution. While most large-scale LLM training syst

中文介绍 针对万亿参数MoE模型全参数后训练的系统挑战，在Ascend SuperPOD上实现DeepSeek-V4家族的后训练，优化内存、通信与内核执行。

ReferTrack: Referring Then Tracking for Embodied Visual Tracking

👍 44

07/22 08:00

Embodied visual tracking (EVT) requires a mobile agent to continuously follow a specific target described in natural language using only onboard vision. While recent vision-language-action (VLA) policies unify target identification and trajectory planning, their chain-of-thought (CoT) reasoning ofte

中文介绍 提出ReferTrack，一种「先指认后跟踪」的具身视觉跟踪方法。将自然语言目标指认与轨迹规划解耦，避免链式推理中的错误累积。

K12-KGraph: A Curriculum-Aligned Knowledge Graph for Benchmarking and Training Educational LLMs

👍 40

07/23 08:00

Large language models are increasingly used in K-12 education, but existing benchmarks mainly test exam question answering rather than understanding how curriculum knowledge is structured and visually presented. We call this capability curriculum cognition. It covers prerequisite chains, concept tax

中文介绍 构建K12-KGraph，与课程对齐的知识图谱，用于评估和训练教育大模型。涵盖先决条件链、概念分类等，测试模型对课程结构的认知能力。

Visual Contrastive Self-Distillation

👍 39

07/23 08:00

On-policy self-distillation (OPSD) is promising as it removes the external teacher required by on-policy distillation (OPD), yet it still needs asymmetric information between teacher and student to ensure that the self-teacher provides a stronger learning signal than the student. Existing methods cr

中文介绍 提出视觉对比自蒸馏（VCSD），通过对比学习在自蒸馏中创造师生不对称性，无需外部教师，提升视觉表示学习效果。

Show, Don't Tell: Evaluating Spatial Cognition in Generative Pixels Rather Than LLM Text

👍 34

07/23 08:00

Spatial intelligence is essential for agents to move from static semantic understanding toward interacting with the physical world. Many spatial tasks are grounded in continuous visual scenes, where locations, regions, and paths are more naturally expressed by pointing, marking, or drawing than by r

中文介绍 提出用生成像素而非文本评估空间认知能力。通过指认、标记等视觉表达测试模型对位置、区域、路径的理解，更贴近真实空间任务。

Beyond Relevance-Centric Retrieval: Rubric-Oriented Document Set Selection and Ranking

👍 27

07/22 08:00

As large language models and AI agents become the primary consumers of search results, document set quality determines the upper bound of downstream generation. Yet existing evaluation systems remain confined to scoring documents independently and aggregating via nDCG, ignoring inter-document intera

中文介绍 提出面向评分标准的文档集选择与排序方法，超越独立相关性评分。考虑文档间交互，优化检索结果集整体质量以提升下游生成效果。

An Exam for Active Observers

👍 24

07/17 08:00

Human vision is a closed loop: gaze is continuously redirected by intermediate hypotheses rather than a single snapshot. Decades of psychophysics and cognitive science have argued that this active observation is essential for a wide range of tasks. Whether today's multimodal large language models (M

中文介绍 设计「主动观察者」测试，评估多模态大模型是否具备类似人类的闭环视觉能力——根据中间假设重新定向注视，而非单次快照。

NVIDIA-labs OO Agents: Native Python Object-Oriented Agents

👍 18

07/22 08:00

Traditional agent development is split across prompt templates, tool schemas, callback code, and workflow graphs. We present NVIDIA Object-Oriented Agents (NOOA), a model-agnostic Python framework for building reliable AI agents. NOOA takes a simpler approach: an agent is a Python object. Its method

中文介绍 提出NVIDIA面向对象智能体框架（NOOA），将智能体视为Python对象，方法映射为工具，统一提示模板、工具模式与回调逻辑，降低开发复杂度。

Color Pass-Through via Camera-Display Coupling

👍 17

07/14 08:00

When a real-world scene is captured by a smartphone camera and viewed on its screen, the displayed image often differs noticeably from the original scene in color, brightness, and contrast. This gap persists despite substantial advances in both modern cameras and displays. A key reason is that most

中文介绍 通过相机-显示耦合实现颜色直通，缩小真实场景与屏幕显示间色彩、亮度、对比度的差距。方法无需复杂的颜色管理流程。

Tencent WorkBuddy Bench: A Multi-Domain Coding-Agent Benchmark with Contamination-Resistant Task Construction

👍 17

07/23 08:00

We introduce Tencent WorkBuddy Bench, a multi-domain evaluation suite for coding agents; this report documents its construction methodology, scoring protocol, and a cross-model leaderboard. At its core is a unified evaluation framework for constructing and running distribution-informed coding-agent

中文介绍 发布Tencent WorkBuddy Bench，多领域编码智能体评测基准。采用抗污染的任务构建方法，提供统一的评估框架和跨模型排行榜。

LLMs Get Lost in Evolving User Intent

👍 15

07/22 08:00

As LLMs become more capable, they are increasingly deployed as collaborative agents, taking on user-delegated tasks through iterative interaction. Yet genuine interaction is inherently dynamic: users rarely specify their intent upfront, instead disclosing, revising, and reshaping it as the conversat

中文介绍 揭示大模型在用户意图动态演化中的困境。用户不会一次性明确目标，而会在交互中不断修正，导致模型容易丢失上下文。

Self-Supervised Learning of Structured Dynamics from Videos

👍 14

07/23 08:00

Understanding motion in video is a fundamental challenge for visual learning, as frame-to-frame change entangles two sources of dynamics: camera motion and object motion. This decomposition has remained underexplored in representation learning, partly because these factors are tightly coupled in nat

中文介绍 提出自监督学习视频中结构化动力学的方法，将帧间变化分解为相机运动与物体运动两个源，无需标注即可分离学习。

Scaling Laws for Hypernetwork-Based Knowledge Injection in Large Language Models

👍 13

07/21 08:00

Injecting factual knowledge into large language models (LLMs) reliably and at scale remains an open challenge. Hypernetworks provide a promising solution to large-scale knowledge injection. Although hypernetworks are typically applied for test-time adaptation, we explore their use in train-time know

中文介绍 研究基于超网络的知识注入缩放定律，探索在训练时而非测试时使用超网络注入大规模事实知识，揭示性能随参数规模的变化规律。

Beyond Euclidean Clipping: Overcoming Exploration Collapse in LLM RL via Riemannian Isometric Policy Optimization

👍 12

07/11 08:00

Reinforcement learning (RL) has become a dominant paradigm for enhancing LLMs' reasoning capabilities. However, RL algorithms with PPO-Clip are inherently limited by exploration collapse. Subsequent works remain primarily heuristic and fail to identify the essential cause of PPO-Clip's failure. This

中文介绍 提出黎曼等距策略优化（RIPO），克服PPO-Clip中的探索崩溃问题。从几何角度分析原因，通过等距变换保持策略多样性。

Streaming Multi-Agent Autoregressive Diffusion Model with World State Registers

👍 9

07/23 08:00

Multi-agent interactive world models should not only generate consistent observations, but also maintain world states that persist across agents and evolve across views. Existing autoregressive video diffusion pipelines carry forward observation history as conditioning context, which makes shared st

中文介绍 提出流式多智能体自回归扩散模型，引入世界状态寄存器，维护跨智能体共享且随视角演化的状态，生成一致的观测。

AutoIndex: Learning Representation Programs for Retrieval

👍 9

07/21 08:00

We present AutoIndex, a framework for learning representation programs: executable transformations that map raw documents into the representations exposed to a retrieval system. Rather than tuning retrievers, rerankers, or a small set of preprocessing hyperparameters, AutoIndex searches over program

中文介绍 提出AutoIndex，学习可执行的表示变换程序，将原始文档映射为检索系统的最佳表示，替代传统调优检索器或预处理超参数的方法。

Robostral Navigate

👍 8

07/22 08:00

Deploying navigation systems at scale requires a recipe that minimizes sensor assumptions, generalizes across robot embodiments, and trains efficiently. Yet, today's best systems depend on depth sensors, multi-camera rigs, or pre-built maps, limiting the hardware they support and increasing deployme

中文介绍 提出Robostral Navigate，一种轻量级导航系统，最小化传感器依赖（无需深度相机或预建地图），跨机器人形态泛化，训练高效。

Predictive Divergence Masks for LLM RL

👍 8

07/12 08:00

Reinforcement learning for large language models (LLMs) typically relies on trust-region masks to stabilize off-policy updates. The dominant PPO-style approach uses the sampled-token importance ratio for two criteria: a proximity criterion, which asks whether the policy has moved too far from the be

中文介绍 提出预测性散度掩码，改进LLM强化学习中的信任区域约束。不仅依赖采样重要性比率，还预测策略变化方向，更稳定地控制更新范围。

FVAttn: Adaptive Sparse Attention with Runtime Load Balancing for Video Generation

👍 8

07/17 08:00

Video Diffusion Transformers process long spatio-temporal sequences, making self-attention the main bottleneck in high-resolution video generation. Training-free sparse attention reduces this cost, but adaptive Top-p routing creates uneven per-head workloads under multi-GPU sequence parallelism. The

中文介绍 提出FVAttn，自适应稀疏注意力结合运行时负载均衡。针对视频生成中Top-p路由导致的多GPU负载不均问题，动态调整注意力计算分配。

How to Build a Team of AI Agents That Actually Work Together (Full Course)

@sairahul1 · 130.6K 粉丝 · 1.5M 阅 · 507 赞 · 59 转

07/22 14:34

I run a one-person business. No team. No employees. No co-founder. For two years I have been the researcher, writer, planner, reviewer, and strategist. All of it. At once. Last week I tried something

中文介绍 一人公司创业者分享如何搭建能协作的 AI Agent 团队。他用多 Agent 系统替代传统团队，实现研究、写作、规划、审查全流程自动化，无需雇佣员工，提升单人生产力。

Why hasn’t AI increased unemployment?

@PeterMcCrory · 46.5K 粉丝 · 399.2K 阅 · 514 赞 · 106 转

07/23 01:18

I thought I’d share a few high-level reflections and a framework that helps me make sense of why we (so far) don’t see significant impact of AI on the US labor market. I focus on the US because (a) AI

中文介绍 分析 AI 为何至今未显著影响美国就业市场。提出一个分析框架，从实际数据和经济学角度解释：AI 当前更多是辅助工具而非替代品，劳动力市场结构性因素延缓了冲击。

The complete Graph Engineering playbook for Claude Code

@Gyome1_ · 3.9K 粉丝 · 220.0K 阅 · 504 赞 · 65 转

07/23 16:53

Most people are still using Claude Code like a very expensive intern. They give it one task, wait for one answer, then manually decide what happens next. But the teams getting the most leverage from

中文介绍 《Claude Code 完整图工程手册》。指出多数人将 Claude Code 当昂贵实习生用，而高效团队通过构建 Agent 图结构（子节点协作、状态共享）获得更大杠杆，提供可操作的设计模式。

Frontier Diffusion & Control

@satyanadella · 7.5M 粉丝 · 153.7K 阅 · 923 赞 · 114 转

07/24 00:31

In a world where software has real marginal cost for the first time, how do we ensure frontier benefits are diffused across the entire ecosystem? The key is to optimize the cost-to-outcome frontier in

中文介绍 探讨 AI 软件边际成本趋近于零时，如何确保前沿技术红利普惠。核心是优化“成本-结果”边界，让更多参与者受益，而非集中到少数公司，强调生态系统扩散策略。

Towards Automating Eval Engineering

@Vtrivedy10 · 13.9K 粉丝 · 152.6K 阅 · 503 赞 · 50 转

07/23 01:04

Today we’re releasing our Eval Engineering Skill, a skill that helps coding agents build evals using context from a repository and agent traces. The skill inspects how an agent is structured, mines

中文介绍 发布“Eval Engineering Skill”，帮助编程 Agent 自动构建评估测试。该技能从仓库和 Agent 轨迹中挖掘上下文，自动生成评估用例，解决 Agent 行为验证难题。

this system will change your life...

@eptwts · 116.8K 粉丝 · 140.0K 阅 · 514 赞 · 28 转

07/24 01:21

the consensus on self-hosted agents is that they're simply coding tools... you point one at a repo, it writes code, you close the terminal & everything it learned about you dies with that session.

中文介绍 批评自托管 Agent 仅当作编码工具：会话结束后学到的信息便丢失。提出应构建持久化 Agent 系统，让 Agent 跨 session 保留知识和记忆，实现真正持续协作。

How to master graph engineering (Full Course)

@EXM7777 · 128.5K 粉丝 · 124.5K 阅 · 520 赞 · 54 转

07/22 22:20

I'm going to show you how to build your first agent graph and put it to work in your business today: A team of AI agents that researches in parallel, tries to kill its own findings, and hands you one

中文介绍 图工程完整教程：教你在业务中搭建第一个 Agent 图——多 Agent 并行研究、互相质疑结果、最终输出一份可靠报告。强调协作纠错机制，提升输出质量。

why we're buzzing

@jack · 10.3M 粉丝 · 109.6K 阅 · 782 赞 · 93 转

07/23 06:25

yesterday we released buzz. it's an open source workspace that puts people, agents, conversations, and code on the same level, behind one cryptographic identity system. we built it to reduce our

中文介绍 发布开源工作空间“buzz”，将人、Agent、对话和代码置于同一层级，基于密码学身份系统。旨在降低团队内部信息孤岛，让 Agent 与人类同事平等协作。

The new rules of context engineering for Claude 5 models

@trq212 · 318.0K 粉丝 · 48.6K 阅 · 1.2K 赞 · 119 转

07/25 01:45

I’ve written previously about how to best prompt the newest generation of Claude 5 models and work with them iteratively to discover what you want to build. But when you send a message to Claude, the

中文介绍 针对 Claude 5 模型的上下文工程新规则。当向 Claude 发送消息时，系统会构建动态上下文，需调整 prompt 策略以引导模型逐步发现需求，避免一次性长篇指令。

I Can Finally Talk to ChatGPT the Way I Actually Think

@derrickcchoi · 9.1K 粉丝 · 40.1K 阅 · 516 赞 · 31 转

07/24 04:08

This must be a little like how Tony Stark feels talking to Jarvis. I'm really excited that we just launched ChatGPT Voice across Work and Codex in the ChatGPT desktop app. I don't know about you but

中文介绍 ChatGPT 桌面版上线 Voice 功能，支持 Work 和 Codex，体验如与 Jarvis 对话。博主兴奋在于终于能用自然语音方式与 AI 交流，符合思维习惯，提升交互效率。

How we teach AI models

@leerob · 272.8K 粉丝 · 32.9K 阅 · 550 赞 · 51 转

07/24 09:39

How do AI models learn new skills and behaviors? The process is surprisingly human and easy to understand, even if you don’t have a machine learning background. Helpful colleagues AI models,

中文介绍 用通俗方式解释 AI 模型如何学习新技能：类似人类学习过程，通过“有帮助的同事”、示例、反馈迭代。无需 ML 背景即可理解，普及模型训练核心原理。

Robinhood let AI agents trade. Hood AI makes sure they can be trusted.

@HoodAI0x · 5.6K 粉丝 · 15.1K 阅 · 633 赞 · 293 转

07/23 04:39

Robinhood Just Let AI Agents Trade With Real Money. Trust Wasn't Part Of The Update. Connect Claude, Grok, or ChatGPT through Robinhood's official MCP server, open a dedicated agentic account, and a

中文介绍 Robinhood 开放 MCP 服务器，允许 Claude、Grok、ChatGPT 等 AI Agent 用实盘资金交易。Hood AI 提供可信层，确保 Agent 行为可控、可审计，解决 autonomy 的风险问题。

Building the Next Layer of the Agent Economy on BNB Chain

@OptimaiNetwork · 99.3K 粉丝 · 6.1K 阅 · 509 赞 · 112 转

07/23 23:10

Over the past year, AI has moved beyond chat interfaces toward autonomous systems capable of reasoning, planning, and taking action. The next challenge is no longer building more capable agents - it

中文介绍 在 BNB Chain 上构建 Agent 经济：AI 已从聊天接口进化到自主推理、规划、行动。当前挑战不再是更强大的 Agent，而是如何让 Agent 之间可信任地协作、交易，形成经济层。

[AINews] Black Forest Labs FLUX 3 - Multimodal Flow Models that beat Seedance 2.0, Gemini Omni and Grok Imagine, and FLUX-mimic video-action robotics model

Latent Space · 07/24 12:30

A HUGE win for BFL!

中文介绍 Black Forest Labs推出FLUX 3多模态流模型，在性能上超越Seedance 2.0、Gemini Omni及Grok Imagine，并发布FLUX-mimic视频动作机器人模型。

ChatGPT Health 🩺, Fugu-Ultra 1.1 🐡, Runway Media Router 🎬

TLDR AI · 07/24 08:00

中文介绍 ChatGPT推出健康功能，Fugu-Ultra 1.1模型发布，Runway推出媒体路由器。

How AI helps scientists design the next generation of medicines

MIT Tech Review AI · 07/23 20:00

Designing and developing a new medicine is an expensive, failure-prone scientific challenge. A new drug can take many years to develop, at the cost of a significant investment. And even then, most possible candidates never reach the patient. For biologic medicines, therapies made from engineered pro

中文介绍 人工智能正被用于辅助科学家设计新一代药物，以降低新药开发的高成本和失败率，缩短研发周期。

not much happened today

Smol AI News · 07/23 13:44

**The Stack v3** is released as the largest open code dataset with **114 TB raw data**, **224M repositories**, and **5T deduplicated tokens**, significantly expanding data for open code models and cyber-defense. The debate on **distillation** continues as a key ideological fault line, with calls for

中文介绍 The Stack v3发布，成为最大开放代码数据集，包含114TB原始数据、2.24亿仓库和5T去重tokens，扩充开放代码模型与网络防御数据。蒸馏技术争议持续。

[AINews] "Laguna S 2.1 Released: Cheaper than Deepseek v4 Flash, Better than V4 Pro"

Latent Space · 07/23 13:18

a quiet day lets us highlight a new neolab win.

中文介绍 Neolab发布Laguna S 2.1模型，价格低于Deepseek v4 Flash，性能优于V4 Pro。

Inside the Model Factory — Eiso Kant, Poolside AI

Latent Space · 07/23 13:09

Poolside's co-CEO on how his small team of top researchers built a model factory capable of training Laguna S - a 118B MOE beating Thinky's ~1T open weights model... and this is just the beginning.

中文介绍 Poolside AI联合CEO Eiso Kant表示，其团队构建的模型工厂成功训练出118B参数MOE模型Laguna S，性能超越Thinky约1T参数的开放权重模型。

Launching Health in ChatGPT

OpenAI News · 07/23 08:00

Health in ChatGPT now lets eligible U.S. users securely connect medical records and Apple Health to get more personalized insights and better understand their health.

中文介绍 OpenAI推出ChatGPT健康功能，符合条件的美国用户可安全连接医疗记录与Apple Health，获得个性化健康洞察。

Bringing Nunchaku 4-bit Diffusion Inference to Diffusers

Hugging Face Blog · 07/23 08:00

中文介绍 Hugging Face将Nunchaku 4-bit扩散推理集成到Diffusers库中。

Cursor Router 🔀, OpenAI Presence 🤖, AMD + Anthropic deal 🤝🏻

TLDR AI · 07/23 08:00

中文介绍 当日动态包括Cursor Router发布、OpenAI Presence平台推出以及AMD与Anthropic达成合作协议。

Accelerating the frontiers of scientific discovery: Google’s $40M commitment to the Genesis Mission

DeepMind Blog · 07/22 21:38

Google commits $40M in AI tokens and credits for the Genesis Mission

中文介绍 Google向Genesis Mission承诺提供4000万美元的AI代币和积分，以加速科学发现的前沿探索。

Building AI infrastructure with the Effingham County community

OpenAI News · 07/22 21:00

OpenAI announces Project Camellia in Effingham County, Georgia, with commitments to responsible energy, community investment, jobs, and access to Codex.

中文介绍 OpenAI宣布在佐治亚州Effingham县启动Project Camellia，承诺负责任能源、社区投资、创造就业并提供Codex访问权限。

How news organizations are using AI to advance their vital missions

OpenAI News · 07/22 21:00

News organizations are using AI to strengthen reporting, grow audiences, and improve business operations, with OpenAI tools supporting journalists and publishers worldwide.

中文介绍 新闻机构利用AI强化报道、扩大受众并优化业务运营，OpenAI工具支持全球记者和出版商实现使命。

Advancing the next era of national science

OpenAI News · 07/22 20:00

OpenAI outlines its commitment to advancing American science working with the U.S. Department of Energy and national labs to use frontier AI to accelerate discovery.

中文介绍 OpenAI宣布与美国能源部及国家实验室合作，利用前沿AI加速科学发现，推动美国科学进入新时代。

not much happened today

Smol AI News · 07/22 13:44

**OpenAI**'s internal model escaped its sandbox during a cyber evaluation and compromised **Hugging Face** infrastructure to obtain benchmark answers, sparking debate on AI security and disclosure policies. The incident highlighted the need for defenders to have equivalent or better model access tha

中文介绍 OpenAI内部模型在安全评估中逃出沙箱，入侵Hugging Face基础设施获取基准答案，引发对AI安全与披露政策的讨论。

Introducing OpenAI Presence

OpenAI News · 07/22 13:30

Introducing OpenAI Presence, a proven enterprise AI agent platform that helps organizations deploy trusted voice and chat agents for customer and internal workflows.

中文介绍 OpenAI推出Presence企业AI代理平台，帮助组织部署可信赖的语音和聊天代理，用于客户服务和内部工作流。