HuggingFace 每日AI论文速递

duan

每天10分钟，带您快速了解当日HuggingFace热门AI论文内容。每个工作日更新，欢迎订阅。 📢播客节目在小宇宙、Apple Podcast平台搜索【HuggingFace 每日AI论文速递】 🖼另外还有图文版，可在小红书搜索并关注【AI速递】

  1. 1d ago

    2026.10.02 | 流式视频主动记忆；音视频扩散奖励路由

    【赞助商】OpenClaw快报每天五分钟，听听 OpenClaw 快报，带你了解最新动态和业内讨论传送门 https://www.xiaoyuzhoufm.com/podcast/6a1732a2dffa135d0ab5ef43 【目录】本期的 15 篇论文如下： [00:29] 🎥 OneStreamer: Unifying Perception, Memory, and Proactive Response in Streaming Video Interaction（OneStreamer：统一流式视频交互中的感知、记忆与主动响应）[01:11] 🔀 Adaptive Reward Routing: Dynamic Multi-Reward Optimization for Joint Audio-Video Diffusion via Forward-Process RL（自适应奖励路由：通过前向过程强化学习实现联合音视频扩散的动态多奖励优化）[01:55] 🧠 Beyond Memory: Harnessing Long-Horizon Agents with Explicit Belief States（超越记忆：利用显式信念状态驾驭长时程智能体）[02:38] 🤖 Agent Priors-guided Policy Learning（智能体先验引导的策略学习）[03:25] 🌀 Hierarchical Continuous Diffusion Language Models（分层连续扩散语言模型）[04:03] 👁 World Observer: Joint Actor-Observer Generation for Persistent World Modeling（World Observer：面向持久世界建模的联合行动者-观察者生成）[04:54] 📉 Sharpening Tax in Post-Training（后训练中的锐化税）[05:39] 🎯 ActiveSaddler: Automated Curriculum Learning for Agent Harness Optimization（ActiveSaddler：面向智能体执行框架优化的自动化课程学习）[06:22] 🤖 A Missing Piece for Trustworthy AI Reviewers: From Benchmarking Rhetorical Robustness to SciCore Review（可信AI审稿人缺失的一环：从修辞鲁棒性基准测试到SciCore评审）[07:08] 🎮 ROWBench: Do Video Models Render What the Program Specifies?（ROWBench：视频模型能否渲染程序所指定的内容？）[07:55] 🤖 AutoGUIWorld: Image Generators as Visual World Models for GUI Agent（AutoGUIWorld：将图像生成器用作 GUI 智能体的视觉世界模型）[08:39] 🔍 Retrieval-Augmented Skill Optimization via Cross-Harness Adaptation（基于跨执行框架适配的检索增强技能优化）[09:23] ⚖ Make Sparse Rewards Count: Density-Aware Reward Aggregation for Multi-Reward RL（让稀疏奖励算数：多奖励强化学习中的密度感知奖励聚合）[10:05] 🤖 Decentralized Master-Mind: Joint Action Refinement through Iterative Intent Denoising in Multi-Agent Pathfinding（去中心化 Master-Mind：多智能体路径规划中通过迭代意图去噪的联合动作精炼）[11:02] 🧩 E-MoE: Enhanced Mixture-of-Experts for Non-Factorized Diffusion Language Models（E-MoE：面向非因子化扩散语言模型的增强混合专家） 【关注我们】您还可以在以下平台找到我们，获得播客内容以外更多信息小红书: AI速递 在小宇宙查看该单集文稿

  2. 2d ago

    2026.10.01 | RIDE外推教师残差；UniEvo-VL在线自蒸馏多模态

    【赞助商】OpenClaw快报每天五分钟，听听 OpenClaw 快报，带你了解最新动态和业内讨论传送门 https://www.xiaoyuzhoufm.com/podcast/6a1732a2dffa135d0ab5ef43 【目录】本期的 15 篇论文如下： [00:27] 🧭 The Teacher Is a Direction, Not a Destination: Extrapolating RL-Induced Representation Residuals in On-Policy Distillation（教师是方向，而非终点：在在线策略蒸馏中外推强化学习诱导的表征残差）[01:11] 🪞 UniEvo-VL: An On-policy Self-Distillation Training Recipe for Multimodal Model Self-improvement（UniEvo-VL：面向多模态模型自我改进的在线策略自蒸馏训练方案）[01:54] 🕵 False Frontiers: Diagnosing and Mitigating Co-Cheating in Self-Evolving Search Agents（虚假前沿：诊断与缓解自演化搜索智能体中的共同作弊）[02:42] 🤖 AREX-2: Advancing Self-Improving Agents through Long-Horizon Reflective Tasks（AREX-2：通过长时程反思任务推进自我改进智能体）[03:29] 🖥 Mid-Harness: Scaling Actions Between Model and Harness for Terminal Agents（Mid-Harness：在模型与执行框架之间扩展终端智能体的动作）[04:09] 🧬 EvoDuet: Bilevel Co-Evolution of Web Searching and Task Solving for Scientific Discovery（EvoDuet：面向科学发现的网络搜索与任务求解双层协同演化）[05:01] 🕵 WorldAuditBench: Interactive 3D World Auditing with Multimodal Agents（WorldAuditBench：使用多模态智能体进行交互式3D世界审计）[05:47] 🛠 Learning Meta-Skills for Agent Harness Design in Test-Time AI4AI（测试时 AI4AI 中面向智能体执行框架设计的元技能学习）[06:26] 🧠 EVOKE: Eliciting World Knowledge in Agents for Transferable Decision-Making（EVOKE：激发智能体中的世界知识以实现可迁移决策）[07:12] 🎮 RSIGame: Autonomous Agentic Game Development with Recursive Self-improvement（RSIGame：具备递归自我改进能力的自主智能体游戏开发）[07:53] 📉 More Choices, Fewer Decisions: Ordinal-Scale Bias in JEV-like Direct-Decision Models（更多选择，更少决策：类JEV直接决策模型中的序数尺度偏差）[08:45] 🧠 Imagine3D-LLM: Teaching MLLMs to Imagine 3D Scenes Before Answering（Imagine3D-LLM：教会多模态大语言模型在回答前想象3D场景）[09:31] 🛠 Agent Error Dataset: Scaling 50,000 Error--Diagnosis Pairs for Failure Analysis and Error-Aware Post-Training（智能体错误数据集：面向失败分析与错误感知后训练，规模化构建5万条错误—诊断配对）[10:21] 🏮 LANTERN: Illuminating Hidden Mathematical Knowledge in Language Models（LANTERN：照亮语言模型中的隐藏数学知识）[11:05] 🖼 It's Not What the Image Shows: Irrelevant Context Destabilises VLM Judges Without Informing Them（并非图像所示：无关上下文会扰乱 VLM 评判模型却不为其提供信息） 【关注我们】您还可以在以下平台找到我们，获得播客内容以外更多信息小红书: AI速递 在小宇宙查看该单集文稿

  3. 2d ago

    【月末特辑】9月最火AI论文 | LimiX-2结构化智能；Vidu S2实时可编辑空间视频

    【赞助商】OpenClaw快报每天五分钟，听听 OpenClaw 快报，带你了解最新动态和业内讨论传送门 https://www.xiaoyuzhoufm.com/podcast/6a1732a2dffa135d0ab5ef43 【目录】本期的 10 篇论文如下： [00:38] TOP1(🔥810) | 🧠 LimiX-2: A Contextual Mechanism Network Towards General Structured-Data Intelligence（LimiX-2：面向通用结构化数据智能的上下文机制网络）[02:36] TOP2(🔥704) | 🎬 Vidu S2: Real-Time Interactive, Editable, and Spatial Video Generation（Vidu S2：实时交互、可编辑与空间视频生成）[04:41] TOP3(🔥494) | 🤖 Raven: The Harness of Harnesses for Composable Agentic Intelligence（Raven：面向可组合智能体智能的“框架之框架”）[06:43] TOP4(🔥493) | 🎓 StudentSim: Training LLM-based Student Simulators（StudentSim：训练基于大语言模型的学生模拟器）[08:51] TOP5(🔥484) | 🤖 Scaling Automatic Research Agents via World Models（通过世界模型扩展自动研究智能体）[11:10] TOP6(🔥431) | 🤖 Atria Dawn: The Dawn of Agentic Superintelligence（Atria Dawn：智能体超级智能的黎明）[13:35] TOP7(🔥396) | 🤖 Repo-To-Skill: Distilling GitHub Repositories Into AI4AI Skills（从仓库到技能：将GitHub代码库蒸馏为AI4AI技能）[15:38] TOP8(🔥382) | 🎨 MaLiang-Harness: A Programmable Path to Image and Video Generation（MaLiang-Harness：通往图像与视频生成的可编程路径）[17:42] TOP9(🔥376) | 🧠 Continual Learning Mechanisms Compose for Long-Horizon Memorization（持续学习机制组合用于长时程记忆）[19:55] TOP10(🔥358) | 🤖 In-Context Learning for Robots: Methods and Applications（面向机器人的上下文学习：方法与应用） 【关注我们】您还可以在以下平台找到我们，获得播客内容以外更多信息小红书: AI速递 在小宇宙查看该单集文稿

  4. 2d ago

    【周末特辑】9月第4周最火AI论文 | 世界模型训练客体永久性；OmniEdu 教学基础模型

    【赞助商】OpenClaw快报每天五分钟，听听 OpenClaw 快报，带你了解最新动态和业内讨论传送门 https://www.xiaoyuzhoufm.com/podcast/6a1732a2dffa135d0ab5ef43 【目录】本期的 5 篇论文如下： [00:48] TOP1(🔥236) | 🧠 Training Object Permanence in World Models（在世界模型中训练客体永久性）[03:10] TOP2(🔥235) | 🎓 OmniEdu: Open Foundation Models for Learning and Teaching（OmniEdu：面向学习与教学的开放基础模型）[05:22] TOP3(🔥218) | 🧬 RRSI: Regularized Recursive Self-Improvement of Agent Harnesses（RRSI：智能体运行框架的正则化递归自我改进）[07:44] TOP4(🔥217) | 🎙 Realtime-Venus: A full-duplex interaction system with asynchronous delegation（Realtime-Venus：一种支持异步委派的全双工交互系统）[09:47] TOP5(🔥160) | 🧭 The Tasteful Agent: Measuring and Improving Taste in Long-Horizon Tasks（有品味的智能体：长时程任务中的品味测量与提升） 【关注我们】您还可以在以下平台找到我们，获得播客内容以外更多信息小红书: AI速递 在小宇宙查看该单集文稿

  5. 5d ago

    2026.09.28 | 层融合正则缩小重建生成差；血缘数据流加速保序容错

    【赞助商】OpenClaw快报每天五分钟，听听 OpenClaw 快报，带你了解最新动态和业内讨论传送门 https://www.xiaoyuzhoufm.com/podcast/6a1732a2dffa135d0ab5ef43 【目录】本期的 15 篇论文如下： [00:28] 🧩 FuseReg: Regularizing Layer Fusion Mitigates the Reconstruction-Generation Gap in Representation Autoencoders（FuseReg：正则化层融合缓解表征自编码器中的重建-生成差距）[01:12] 🔗 RayOrch: Programming and Executing Lineage-Controlled Multi-Grain Dataflows for Foundation-Model Data Preparation（RayOrch：面向基础模型数据准备的受血缘控制多粒度数据流编程与执行）[01:58] ⚡ Block Sparse Attention with Log-Linear Complexity（对数线性复杂度的块稀疏注意力）[02:40] 🤖 InternW0-$Δ$: A World Action Model Bridging Predictive Dynamics and Actions with 20K+ Hours of Open Data（InternW0-Δ：一个连接预测动态与动作、基于20K+小时开放数据的世界动作模型）[03:24] 🖐 Tactile-JEPA: Topology-Aware Self-Supervised Representation Learning for Distributed Tactile Sensors（Tactile-JEPA：面向分布式触觉传感器的拓扑感知自监督表示学习）[04:11] 🛰 Enhancing Photogrammetric Digital Surface Models with Pretrained Diffusion Models and Multimodal Conditioning（利用预训练扩散模型与多模态条件增强摄影测量数字表面模型）[04:58] 👁 FoMo: Forking Moment in Generative Trajectory as a Perceptual Distance（FoMo：生成轨迹中的分叉时刻作为感知距离）[05:42] 📊 Jev in the Wild: A Data-Driven Analysis of the Jev Model's Functionality, Applications and Ecosystem（真实环境中的 Jev：Jev 模型功能、应用与生态的数据驱动分析）[06:30] 🎯 TrackEverything: Long Horizon Dense Tracking via De-Duplicating 3D Scene Representations（TrackEverything：通过去重3D场景表示实现长时程密集跟踪）[07:21] 🧩 SLCA-GRPO: Resolving Cross-Segment Credit Misattribution in Tool-Calling RL（SLCA-GRPO：解决工具调用强化学习中的跨段信用误归因）[08:08] 🎯 CARD: Cluster-level Adaptation with Reward-guided Decoding for Personalized Text Generation（CARD：面向个性化文本生成的聚类级适配与奖励引导解码）[08:54] ⚖ Do Implicit Personalization and Explicit Styles Conflict? PsPLUG: A Lightweight Plug-in for Balancing Personalization and Style in Customized LLMs（隐式个性化与显式风格会冲突吗？PsPLUG：用于平衡定制化 LLM 中个性化与风格的轻量级插件）[09:41] 🤝 AgentWorld: Benchmarking Long-Horizon Collaboration of Multi-agent LLMs（AgentWorld：多智能体大语言模型长时程协作基准测试）[10:24] 🎮 Game Arena: Strategic LLM Evaluation in Competitive Environments（游戏竞技场：竞争环境中的大语言模型策略评估）[11:08] 🏦 IndicBankBench: Evaluating Safety and Reliability of Language Model Assistants in Indian Retail Banking（IndicBankBench：评估印度零售银行中语言模型助手的安全性与可靠性） 【关注我们】您还可以在以下平台找到我们，获得播客内容以外更多信息小红书: AI速递 在小宇宙查看该单集文稿

  6. Sep 26

    2026.09.25 | 世界模型物理推理可评测；大模型线性叠加双想法

    【赞助商】OpenClaw快报每天五分钟，听听 OpenClaw 快报，带你了解最新动态和业内讨论传送门 https://www.xiaoyuzhoufm.com/podcast/6a1732a2dffa135d0ab5ef43 【目录】本期的 15 篇论文如下： [00:27] 🧠 Training Object Permanence in World Models（在世界模型中训练客体永久性）[01:12] 🧠 Your Transformer Can Hold Two Thoughts at Once: Evidence of Linear Superposition in LLMs（你的Transformer能同时容纳两个想法：LLM中线性叠加的证据）[01:53] 🎬 WanPE: Towards Cinematic Prompt Enhancement for Modern Text-to-Video Generation（WanPE：面向现代文本到视频生成的电影级提示增强）[02:40] 🧭 OmniEcho: Audio-Visual Spatial Understanding for Omni-Modal Embodied Agents（OmniEcho：面向全模态具身智能体的音视频空间理解）[03:18] 🤖 Agent-Editing World Model: Rethinking World Modeling for LLM Agents（智能体编辑世界模型：为LLM智能体重新思考世界建模）[04:05] 🧩 Parts-of-Speech as Emergent Categories in SAE Latent Space（词性作为SAE潜空间中的涌现类别）[04:46] 🤖 Qwen-Planner-Agent: A Closed-Loop AI-for-AI Framework for Real-World Mobile Planner Agents（Qwen-Planner-Agent：面向真实世界移动规划智能体的闭环 AI-for-AI 框架）[05:20] 🔍 IterSynth: Rethinking Deep Search Agents via Role-Decoupled Iterative Synthesis（IterSynth：通过角色解耦的迭代合成重新思考深度搜索智能体）[06:04] 🤖 Coding Agents for Generalized Task and Motion Planning Problems（面向泛化任务与运动规划问题的编码智能体）[06:49] 🧠 Neural Spectral Capacity: Measuring and Designing Architectures from Network Specification Alone（神经谱容量：仅凭网络规格测量与设计架构）[07:28] 🛡 AgentKernel: The Trust-Native Agentic Operating System（AgentKernel：信任原生的智能体操作系统）[08:15] 🧪 Rufus-Air: An Open LLM Post-Training Recipe（Rufus-Air：开放的大语言模型后训练配方）[08:59] 🤖 PUBG Ally: A Conversational Embodied Agent as an AI Teammate（PUBG Ally：作为AI队友的对话式具身智能体）[09:44] 🤖 World Action Agent: Harnessing VLMs for Robot Manipulation via World Action Rehearsal（世界动作智能体：利用视觉语言模型通过世界动作预演实现机器人操作）[10:27] 🛸 ExplorationBench: Measuring AI Systems' Exploration in Verifiable Alien Worlds（ExplorationBench：测量 AI 系统在可验证异星世界中的探索能力） 【关注我们】您还可以在以下平台找到我们，获得播客内容以外更多信息小红书: AI速递 在小宇宙查看该单集文稿

  7. Sep 25

    2026.09.24 | 说话者双轨记忆；交互学习空间推理

    【赞助商】OpenClaw快报每天五分钟，听听 OpenClaw 快报，带你了解最新动态和业内讨论传送门 https://www.xiaoyuzhoufm.com/podcast/6a1732a2dffa135d0ab5ef43 【目录】本期的 15 篇论文如下： [00:28] 🧠 SpeakerMem-R1: Speaker-Centered Dual-Track Memory for Multi-Party Dialogue（SpeakerMem-R1：以说话者为中心的多方对话双轨记忆）[01:13] 🤖 Spatial-Interactor: Learning Spatial Reasoning through Interaction with the Observable Physical World（Spatial-Interactor：通过与可观测物理世界交互学习空间推理）[01:59] 🌍 HappyWorld-Bench（快乐世界基准（HappyWorld-Bench））[02:40] 🧠 The Past Frames the Future: Memory for Autoregressive Video Generation（过往帧塑造未来：自回归视频生成中的记忆机制）[03:29] 🧠 Just-in-Time Memory: Learning to Curate Task-Adaptive Memory for LLM Agents（即时记忆：学习为LLM智能体整理任务自适应记忆）[04:11] 🎬 RewardVerse: Rubric-Guided Policy Optimization for Video Reward Modeling（RewardVerse：面向视频奖励建模的评分量规引导策略优化）[05:00] 🎯 PACT: From Credit Assignment to Critic Alignment（PACT：从信用分配到评论家对齐）[05:43] 🧪 Schrödinger's Code Repository: Have LLMs Learned SWE-bench or Memorized It?（薛定谔的代码仓库：LLM 是学会了 SWE-bench，还是记住了它？）[06:24] 📐 GeoPair: Geometry-Preserving Cross-Layer Factorization for Training-Free Transformer Compression（GeoPair：用于免训练 Transformer 压缩的几何保持跨层因子分解）[07:00] 📦 PackLab: A Comprehensive Framework for Developing, Training, and Evaluating MLLMs in Robotic Bin Packing（PackLab：在机器人装箱中开发、训练与评估多模态大语言模型的综合框架）[07:53] 🧠 MemBodied: Recurrent Associative Memory for Vision-Language-Action Models（MemBodied：面向视觉-语言-动作模型的循环联想记忆）[08:39] 🧪 WhatWorkedBench: Benchmarking Experimental Understanding in AI Agents（WhatWorkedBench：基准测试AI智能体的实验理解能力）[09:21] 🧠 Hunyuan-A13B Technical Report（混元-A13B 技术报告）[10:03] 🎮 Verifiable Hidden Dynamics Play: Generating Agentic RL Environments from Solved Mechanisms（可验证隐藏动力学游戏：从已求解机制生成智能体强化学习环境）[10:45] 🎥 All modalities are equal, but video is more equal: Closing the Cross-Attention Gap in Joint Video Generation（所有模态都是平等的，但视频更平等：弥合联合视频生成中的交叉注意力差距） 【关注我们】您还可以在以下平台找到我们，获得播客内容以外更多信息小红书: AI速递 在小宇宙查看该单集文稿

  8. Sep 22

    2026.09.22 | VLM智能迁移至机器人控制；隐式3D记忆构建视频世界模型

    【赞助商】OpenClaw快报每天五分钟，听听 OpenClaw 快报，带你了解最新动态和业内讨论传送门 https://www.xiaoyuzhoufm.com/podcast/6a1732a2dffa135d0ab5ef43 【目录】本期的 15 篇论文如下： [00:28] 🤖 Transferring the Intelligence of VLMs to Robotic Control（将VLM的智能迁移至机器人控制）[01:11] 🎥 WorldCrafter: Consistent Video World Model with Implicit 3D-aware Memory（WorldCrafter：具有隐式3D感知记忆的一致视频世界模型）[01:53] 🎮 GameHorizon Suite: Multi-Horizon Data and Evaluation in Gameplay（GameHorizon 套件：游戏玩法中的多时间跨度数据与评估）[02:35] 🧬 RRSI: Regularized Recursive Self-Improvement of Agent Harnesses（RRSI：智能体运行框架的正则化递归自我改进）[03:20] 📄 Document Retrieval-Aware Chunking (D-RAC): Universal Retrieval-Aware Ingestion of Enterprise Documents via PDF Normalization and Multimodal Markdown Conversion（文档检索感知分块（D-RAC）：通过 PDF 规范化与多模态 Markdown 转换实现企业文档的通用检索感知摄取）[04:01] 🎓 OmniEdu: Open Foundation Models for Learning and Teaching（OmniEdu：面向学习与教学的开放基础模型）[04:47] 🎬 VideoGen-Agent: Reinforcing Video Generation Agents（VideoGen-Agent：强化视频生成智能体）[05:37] 🐼 onPanda: Efficient Annotation of On-Policy Alignment Data for LLMs and Agents via Token-Level Correction（onPanda：通过Token级纠正为LLM与智能体高效标注同策略对齐数据）[06:19] 🤖 Grounded Action Model: 3D Grounding as a Foundation for Robotics（接地动作模型：以3D接地作为机器人学基础）[07:05] 🤖 One to More, More to One: Category-Aware Iterative Expert Training for Software Engineering Agents（从一到多，从多到一：面向软件工程智能体的类别感知迭代专家训练）[07:57] 🎭 Deep Persona: A Psychologically Grounded Architecture and Evaluation Framework for Role-Playing Agents and Simulations（Deep Persona：面向角色扮演智能体与模拟的心理学基础架构与评估框架）[08:36] 🧠 Harness-Zero: Harness Distillation via Agent-as-Harness（Harness-Zero：通过智能体作为外壳进行外壳蒸馏）[09:18] 🤖 CARE: Experience-Guided Atomic Corrective Execution for Vision-Language-Action Policies（CARE：面向视觉-语言-动作策略的经验引导式原子纠正执行）[10:13] 🧠 Jev-Mem: System-One-Controlled Agentic Memory for Efficient AI Agents（Jev-Mem：面向高效 AI 智能体的 System-One 控制型智能体记忆）[10:52] 🎥 Why Do Video Diffusion Models Violate Physics? Unveiling the Flaws in Attention Mechanisms（为什么视频扩散模型会违背物理规律？揭示注意力机制中的缺陷） 【关注我们】您还可以在以下平台找到我们，获得播客内容以外更多信息小红书: AI速递 在小宇宙查看该单集文稿

See All (713)

Ratings & Reviews

5
out of 5
2 Ratings

About

每天10分钟，带您快速了解当日HuggingFace热门AI论文内容。每个工作日更新，欢迎订阅。 📢播客节目在小宇宙、Apple Podcast平台搜索【HuggingFace 每日AI论文速递】 🖼另外还有图文版，可在小红书搜索并关注【AI速递】

Information

You Might Also Like