2026-07-13

Accurate, Interdisciplinary and Transparent Structure-property Understanding with Deep Native Structural Reasoning

👍 84

07/08 08:00

Structure-property relationships are foundational to biology, chemistry and materials science, where function, reactivity and physical response emerge from spatial, chemical and periodic organization. Mechanistically explaining these relationships requires interpreting structural evidence through sc

中文介绍 提出Deep Native Structural Reasoning方法，直接对原始结构证据进行深度推理，实现材料、化学、生物领域结构-性质关系的准确、跨学科、透明理解，无需人工特征工程，提升可解释性与泛化能力。

Video-Oasis: Rethinking Evaluation of Video Understanding

👍 54

07/02 08:00

The inherent complexity of video understanding makes it difficult to determine whether Video-LLM benchmark performance stems from visual perception, linguistic reasoning, or knowledge priors. While many benchmarks have emerged to assess high-level reasoning, shared criteria for evaluating video unde

中文介绍 Video-Oasis重新审视视频理解评估，指出当前基准难以区分性能来源是视觉感知、语言推理还是知识先验，提出共享评价标准以更公平衡量Video-LLM的真实理解能力。

Dual Latent Memory in Vision-Language-Action Models for Robotic Manipulation

👍 51

07/08 08:00

Mainstream Vision-Language-Action (VLA) models predict actions primarily from the current observation under a Markovian assumption, thus struggling with long-horizon, temporally dependent tasks. Existing memory-augmented VLAs either expand the observation window or retrieve history from the memory b

中文介绍 针对VLA模型在长时任务中的马尔可夫假设局限，提出Dual Latent Memory，融合扩展观察窗口与历史检索，增强机器人操控对时间依赖任务的记忆与推理能力。

Infinite Worlds with Versatile Interactions

👍 31

07/08 08:00

We present LingBot-World 2.0 (also known as LingBot-World-Infinity), an advanced iteration of LingBot-World featuring four distinct upgrades. (1) Our model achieves an unbounded interaction horizon while maintaining consistent output quality, benefiting from a carefully crafted causal pretraining pa

中文介绍 LingBot-World 2.0实现无限交互世界，通过因果预训练范式在无限长交互中保持输出一致性，支持多样化场景生成，适用于大规模虚拟环境与交互仿真。

Ideas Have Genomes: Benchmarking Scientific Lineage Reasoning and Lineage-Grounded Idea Generation

👍 30

07/09 08:00

Scientific ideas rarely start from a blank page. They inherit mechanisms, repair known limitations, and recombine pieces of earlier work, much like biological genomes. Current benchmarks still say little about whether AI systems can follow this inheritance structure. We present IdeaGene-Bench (IG-Be

中文介绍 提出IdeaGene-Bench基准，评估AI系统追踪科学思想遗传谱系的能力，要求模型基于前人工作推理并生成有谱系依据的新想法，衡量科学推理深度。

LongE2V: Long-Horizon Event-based Video Reconstruction, Prediction, and Frame Interpolation with Video Diffusion Models

👍 27

07/09 08:00

Recovering high-quality video from sparse event streams is a challenging task. Regression methods often blur textures, while existing generative models struggle with long-term stability. We propose LongE2V, a novel approach that leverages pre-trained video diffusion priors to jointly handle event-ba

中文介绍 LongE2V利用预训练视频扩散先验，联合处理事件流视频重建、预测与帧插值，在长时间跨度下保持稳定，避免回归方法的纹理模糊，实现高质量事件视频恢复。

UniClawBench: A Universal Benchmark for Proactive Agents on Real-World Tasks

👍 27

07/09 08:00

The rapid development of large language models and multimodal large language models has accelerated the emergence of proactive agents capable of operating everyday tools and assisting users in real-world environments. However, existing benchmarks struggle to evaluate such agents effectively, as they

中文介绍 UniClawBench提供通用基准，评估能操作工具、协助用户的主动智能体在真实世界任务中的表现，弥补现有基准在任务多样性和环境真实性方面的不足。

LLM-as-a-Tutor: Policy-Aware Prompt Adaptation for Non-Verifiable RL

👍 27

07/05 08:00

Reinforcement learning (RL) for non-verifiable instruction following increasingly relies on LLM judges with prompt-specific rubrics as reward signals. While recent methods adapt these rubrics to the evolving policy during training, the training prompts themselves remain static, drawn from fixed corp

中文介绍 提出策略感知提示自适应方法，在非可验证指令跟随的RL中，根据策略演化动态调整LLM评判的评分细则与训练提示，提升奖励信号的准确性与对齐效果。

Enhancing In-context Panoramic Generation via Geometric-aware Pretraining

👍 17

07/09 08:00

In this work, we present Canvas360, a two-stage framework for in-context panoramic generation that combines geometry-aware pretraining with downstream task-specific fine-tuning. To address the lack of large-scale, high-quality training data tailored to in-context panoramic tasks, we propose Canvas36

中文介绍 Canvas360两阶段框架：先通过几何感知预训练获得大尺度全景数据表征，再微调实现上下文全景生成，有效缓解训练数据不足问题，提升生成质量。

Jet-Long: Efficient Long-Context Extension with Dynamic Bifocal RoPE

👍 15

07/08 08:00

Modern LLMs are increasingly deployed in long-context applications such as retrieval-augmented generation, repository-level coding, and agentic workflows whose accumulated reasoning and tool traces routinely push the input an order of magnitude past the pretraining window, making zero-shot context e

中文介绍 Jet-Long提出动态双焦点RoPE，高效扩展LLM上下文窗口至远超预训练长度，通过聚焦重要位置与稀疏化注意力，在RAG、代码仓库等任务中实现零样本长上下文扩展。

JD Oxygen AI Item Center (Oxygen AIIC) V1: An Industrial-Scale LLM/VLM-Centric Solution for Item Understanding, Management, and Applications

👍 15

06/29 08:00

JD.com, one of the world's largest e-commerce platforms, serves over 700 million active users and millions of merchants, with a catalog of tens of billions of SKUs. At this scale, high-quality, structured item knowledge underpins a better consumer experience, lower management costs, and higher opera

中文介绍 京东推出Oxygen AIIC V1，以LLM/VLM为核心的工业级商品理解管理系统，在数十亿SKU规模下自动提取结构化商品知识，提升用户体验与管理效率、降低成本。

Single-Rollout Asynchronous Optimization for Agentic Reinforcement Learning

👍 14

07/08 08:00

Reinforcement learning (RL) is becoming increasingly important for post-training large language models (LLMs). Previous RL pipelines for LLMs were mostly synchronous and batch-interleaved, which is inefficient for long-horizon agentic tasks. Recently, asynchronous RL has emerged as a more efficient

中文介绍 提出单轨迹异步优化用于智能体强化学习，相比同步批处理方法，在长程智能体任务中显著提升训练效率，适用于LLM后训练阶段的agentic RL场景。

OpenCoF: Learning to Reason Through Video Generation

👍 13

07/09 08:00

Reasoning has become a core capability for large models, especially when reliable decisions require understanding logical consequences. Recent video generation models offer a reasoning path distinct from previous Chain-of-Thought (CoT): reasoning can unfold through temporally connected frames, known

中文介绍 OpenCoF通过视频生成实现推理，推理过程在时序帧中展开，模型学习生成因果连贯的视频帧序列，以视觉方式呈现逻辑推理，提升可解释性。

CineMobile: On-Device Image-to-Video Diffusion for Cinematic Camera Motion Generation

👍 12

07/04 08:00

The growing demand for image-to-video creation on mobile devices has increasingly focused on cinematic motion effects like bullet time, dolly zoom, slow motion, etc. While Diffusion Transformers (DiTs) exhibit strong performance in video generation, their large parameter sizes and multi-step iterati

中文介绍 CineMobile面向移动端图像生成视频，实现子弹时间、滑动变焦等电影级运镜，通过轻量Diffusion Transformer与多步迭代优化在设备上高效生成高质量动态效果。

RoboDojo: A Unified Sim-and-Real Benchmark for Comprehensive Evaluation of Generalist Robot Manipulation Policies

👍 11

07/07 08:00

Generalist robot manipulation policies have advanced rapidly, yet existing benchmarks remain limited in systematically evaluating their capabilities. Many rely on simple, short-horizon, or skill-narrow tasks with limited capability coverage, and are often conducted only in simulation or only in the

中文介绍 RoboDojo提供统一的仿真与真实基准，系统评估通用机器人操控策略，覆盖多技能、长时域任务，在两种环境中全面衡量泛化能力，弥补现有基准局限。

Remember When It Matters: Proactive Memory Agent for Long-Horizon Agents

👍 10

07/09 08:00

In long-horizon tasks, decision-relevant state is often scattered across an expanding trajectory, while the action agent must surface it and act. As trajectories grow, task requirements, environment facts, prior attempts, diagnoses, and open subgoals can be buried in the context window or pushed bey

中文介绍 提出Proactive Memory Agent，主动管理长时任务中的决策相关状态，通过记忆检索与上下文整合防止关键信息被长轨迹淹没，提升智能体在复杂长程任务中的表现。

Sparse Delta Memory: Scaling the State of Linear RNNs through Sparsity

👍 10

07/08 08:00

Linear attention models allow a fixed state size and a fixed amount of compute per token. However, due to their limited state size, linear attention models fall behind in long-context recall compared to softmax-attention-based transformer architectures. Increasing the state size of linear attention

中文介绍 Sparse Delta Memory利用稀疏机制扩展线性RNN的状态容量，在不增加计算量的前提下提升长上下文召回能力，弥合线性注意力与softmax注意力在长程记忆上的差距。

Automating the Design of Embodied Agent Architectures

👍 10

07/03 08:00

Embodied agents are typically built as hand-designed compositions of perception, memory, planning, and action modules. This modularity exposes a large architectural design space, but current systems still rely on researcher intuition to choose where information is stored, how observations are proces

中文介绍 提出自动化设计具身智能体架构的方法，搜索感知、记忆、规划、行动等模块的最优组合，替代手工设计，自动探索更大架构空间，提升任务适应性与性能。

WildCity: A Real-World City-Scale Testbed for Rendering, Simulation, and Spatial Intelligence

👍 9

07/07 08:00

Humans can navigate an unfamiliar city and gradually form a coherent spatial mental map spanning tens of square kilometers. Can AI build spatial representations at a comparable scale? Although recent foundation models have advanced scene reconstruction and embodied intelligence, scaling to entire ci

中文介绍 WildCity提供城市尺度的真实世界测试台，覆盖数平方公里场景，支持渲染、仿真与空间智能研究，用于评估AI在大型陌生环境中构建空间表征、导航与推理的能力。

Attending to Multimodal Generation One Token at a Time

👍 8

07/04 08:00

Multimodal large language models (MLLMs) generate responses autoregressively, integrating visual and linguistic information in an evolving context. Prior work on interpretability has focused on individual layers and circuits (where), leaving the token-level dynamics of multimodal computation during

中文介绍 研究MLLM在自回归生成中每个token的多模态计算动态，通过细粒度注意力分析揭示视觉与语言信息逐步整合的token级机制，深化对多模态生成过程的理解。

Copy Claude Fable 5’s Thinking Before It’s Gone

@TheAIColony · 225.1K 粉丝 · 1.2M 阅 · 603 赞 · 140 转

07/10 15:57

You Have a Few Days to Copy Fable 5’s Brain Into a Cheaper Model. Here Is How. July 12 is the last day Claude Fable 5 sits inside your plan for free. After that, it moves to pay-per-use credits, and

中文介绍 Claude Fable 5 即将于7月12日转为按量付费，博主提醒用户在此之前将其思维模式迁移到更便宜的模型，并提供了操作方法。

The Great Wave Has Arrived (from GLM CEO Jie Tang)

@bingxu_ · 6.5K 粉丝 · 318.6K 阅 · 505 赞 · 88 转

07/11 23:10

-- Bing Xu's Note --- I came across an internal GLM letter on the Chinese app RedNote, purportedly written by @jietang, and translated the Chinese text in the images into English. --- Letter Start

中文介绍 翻译了GLM CEO Jie Tang 在 RedNote 发布的内部信《大浪潮已至》，内容涉及AI趋势，可能引发行业讨论。

The Reverse Information Paradox

@satyanadella · 7.0M 粉丝 · 146.5K 阅 · 1.1K 赞 · 162 转

07/12 23:09

In the age of intelligence, how should firms protect their core IP? Nobel Prize winning economist Kenneth Arrow famously described a paradox in the market for information. “Its value for the

中文介绍 探讨AI时代企业保护核心IP的困境，引用Arrow信息悖论，指出信息价值在不确定市场中难以评估，需反向思考策略。

Cost effective harnesses with Fable

@RLanceMartin · 36.7K 粉丝 · 119.8K 阅 · 510 赞 · 51 转

07/11 02:00

There is a lot of interest in cost effective use of Fable 5. Agent harnesses will get better at knowing when to use frontier intelligence. I wanted to share some tests I’ve run to better understand

中文介绍 分享经济使用Fable 5的测试结果：Agent harnesses 能智能判断何时调用前沿模型，从而降低推理成本，实现更经济有效的使用。

ORACLE: Official AI Agents Trade on Polymarket

@Oracle__Promo · 4.1K 粉丝 · 106.5K 阅 · 1.8K 赞 · 543 转

07/11 07:04

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 介绍ORACLE：官方AI Agent 在预测市场Polymarket自动交易，数据显示超30%活动来自AI钱包，证明其策略有效性。

ORACLE: Official AI Agents Trade on Polymarket

@Oracle__Trade · 4.0K 粉丝 · 105.3K 阅 · 1.9K 赞 · 557 转

07/11 07:03

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 ORACLE官方AI Agent在Polymarket交易，超过30%活动来自AI钱包，展示自主交易策略的高效性。

gpt-5.6-sol without hitting limits

@theo · 356.6K 粉丝 · 77.0K 阅 · 774 赞 · 27 转

07/12 06:59

I've burned over $200,000 of tokens with gpt-5.6-sol. It's a great model. That said, it's a bit too easy to hit your limits on the $200 Codex Pro sub. OpenAI has been generous about resets, but that

中文介绍 作者使用gpt-5.6-sol消耗超20万美元token，但频繁触达$200 Codex Pro订阅限制，尽管OpenAI会重置，仍需注意配额问题。

GPT‑5.6 by our community

@OpenAIDevs · 373.8K 粉丝 · 58.5K 阅 · 532 赞 · 34 转

07/12 03:14

GPT‑5.6 has only been in builders’ hands for a couple of days. Here’s what the community is already building with it: A voxel Manhattan, built autonomously with GPT‑5.6 Sol: GPT‑5.6 Sol got to

中文介绍 GPT-5.6刚发布几天，社区已用它自主构建体素曼哈顿城市等项目，展示强大的生成能力和社区创造力。

A single question to track progress from o3 to gpt-5.6 and beyond

@SebastienBubeck · 78.1K 粉丝 · 47.8K 阅 · 514 赞 · 42 转

07/10 23:04

How long can be the path of a gradient flow on a convex function given the constraint that it stays within the unit Euclidean ball in dimension n? ^ this is the question I have been using for 2 years

中文介绍 用一个数学问题（凸函数上梯度流路径长度）持续两年追踪从o3到GPT-5.6的模型能力进步，作为基准测试。

[AINews] not much happened today

Latent Space · 07/11 10:53

a quiet day after a week of nonstop model releases

中文介绍 平静的一天，上周连续发布模型后无重大事件。

How Deutsche Telekom is rewiring telecommunications with AI

OpenAI News · 07/10 15:00

How Deutsche Telekom is becoming an AI-native telco with OpenAI-transforming customer service, employee workflows, network operations, and the future of voice.

中文介绍 德国电信与OpenAI合作，用AI转型客户服务、员工工作流、网络运营和语音未来，成为AI原生电信公司。

[AINews] OpenAI launches GPT 5.6 Sol/Terra/Luna, Codex becomes ChatGPT superapp

Latent Space · 07/10 14:19

A big day for OpenAI.

中文介绍 OpenAI发布GPT-5.6系列（Sol/Terra/Luna），Codex升级为ChatGPT超级应用。

not much happened today

Smol AI News · 07/10 13:44

**OpenAI** rolled out **GPT-5.6** featuring a new model stratification with tiers **Luna / Terra / Sol** and effort levels including **Max** and **Ultra**, introducing complex configuration options. The launch faced UX challenges with the **ChatGPT Work / Codex** split, prompting rapid corrective ac

中文介绍 OpenAI推出GPT-5.6，模型分层为Luna/Terra/Sol，引入Max和Ultra努力级别，配置复杂；同时ChatGPT Work与Codex的分裂带来UX挑战。

Profiling in PyTorch (Part 3): Attention is all you profile

Hugging Face Blog · 07/10 08:00

中文介绍 PyTorch性能分析系列第三篇，主要讨论注意力机制的性能剖析。

GPT-5.6 🚀, Muse Spark 1.1 ✨, ChatGPT Work 💼

TLDR AI · 07/10 08:00

中文介绍 当日AI新闻：GPT-5.6发布，Muse Spark 1.1更新，ChatGPT Work上线。

Anthropic found a hidden space where Claude puzzles over concepts

MIT Tech Review AI · 07/10 04:22

The AI firm Anthropic has developed a technique that has given it the clearest glimpse yet at what’s really going on inside large language models as they answer questions or carry out tasks. What they found ranges from the mundane to the unnerving. Researchers at the company built a tool called the

中文介绍 Anthropic开发新技术，揭示了大模型在回答问题或执行任务时的内部运作，发现结果从平凡到令人不安。

GPT-5.6 is now the preferred model in Microsoft 365 Copilot

OpenAI News · 07/09 21:00

Learn how GPT-5.6 powers Microsoft 365 Copilot with stronger AI capabilities across Word, Excel, PowerPoint, Chat, and Cowork for faster, higher-quality work.

中文介绍 GPT-5.6成为Microsoft 365 Copilot首选模型，增强Word、Excel、PowerPoint、Chat和Cowork的AI能力，实现更快更高质量的工作。

GPT-5.6: Frontier intelligence that scales with your ambition

OpenAI News · 07/09 18:00

More intelligence from every token, stronger performance per dollar, and more capability on demand for your hardest work.

中文介绍 GPT-5.6提供每个token更多智能、更强的每美元性能，以及按需扩展的能力，应对最具挑战的工作。

ChatGPT is now a partner for your most ambitious work

OpenAI News · 07/09 18:00

ChatGPT Work is an agent that can take action across your apps and files, stay with a project for hours if needed, and turn a goal into finished work.

中文介绍 ChatGPT Work作为智能代理，能跨应用和文件执行操作，长时间跟踪项目，将目标转化为完成的工作。

GPT-5.5 Bio Bug Bounty

OpenAI News · 07/09 18:00

Details about the OpenAI Bio Bounty program

中文介绍 OpenAI推出GPT-5.5生物漏洞赏金计划，旨在发现生物相关风险。

[AINews] SpaceXAI launches Grok 4.5, first Opus-class model post Cursor acquisition

Latent Space · 07/09 14:05

SpaceXAI continues to move faster than any other frontier lab on earth.

中文介绍 SpaceXAI发布Grok 4.5，是收购Cursor后首个Opus级模型，速度领先其他前沿实验室。

OpenAI launches GPT 5.6 Sol/Terra/Luna

Smol AI News · 07/09 13:44

**OpenAI** launched the **GPT-5.6** family with three models: **Sol**, **Terra**, and **Luna**, integrated across **ChatGPT**, **Codex**, and the API. Pricing tiers range from **$1 to $5 per million tokens** with new cache-write pricing and a 90% cache-read discount. The launch includes new app feat

中文介绍 OpenAI发布GPT-5.6家族（Sol/Terra/Luna），集成于ChatGPT、Codex和API；定价每百万token 1-5美元，提供缓存写入定价和90%缓存读取折扣。

Grok 4.5 🤖, GPT-Live 🎙️, SWE-1.7 👨‍💻

TLDR AI · 07/09 08:00

中文介绍 当日AI新闻：SpaceXAI发布Grok 4.5，OpenAI推出GPT-Live语音功能，以及SWE-1.7编码助手。

Why AI Infrastructure must evolve for Agent Experience — Akshat Bubna, Modal CTO

Latent Space · 07/09 06:55

2 years after our first coverage, we return with Modal's other cofounder to explore why Agent Experience is working now, and everything they have learned building the new agent cloud.

中文介绍 Modal CTO Akshat Bubna探讨AI基础设施需为Agent体验演进，分享构建新一代代理云的经验。