HuggingFace 每日AI论文速递

【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：[00:31] 🤖 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）[01:26] 🌍 SolarWM: Open Data and Scalable Training for Long-Horizon Video World Models（SolarWM：面向长时程视频世界模型的开放数据与可扩展训练）[02:19] ⏱ EarlyEval: Cheaper Agent Evaluation via Early Outcome Prediction（EarlyEval：通过早期结果预测实现更廉价的智能体评估）[03:21] 🤝 It Takes Two to Match: Co-Evolving Generative Retriever with Reinforcement Learning（两相匹配：基于强化学习的生成式检索器协同进化）[04:18] 🎯 Language Models Can Control Their Own Attention（语言模型能够控制自身的注意力）[05:10] 🖼 On the Design Fundamentals of Pixel Text Representation Learning（论像素文本表示学习的设计基础）[06:12] 🔍 Beyond Visual Similarity: Entity-Aligned Retrieval for Knowledge-Based Visual Question Answering（超越视觉相似性：面向知识型视觉问答的实体对齐检索）[07:14] 🤖 HarnessDev: Can LLMs Create and Evolve Their Own Agent Harness?（HarnessDev：大语言模型能否创建并演化自己的智能体执行框架？）[08:05] 🗜 ZipTok3D: High-Fidelity 3D Tokenization with Compact Token Prefixes（ZipTok3D：利用紧凑令牌前缀的高保真三维令牌化）[09:04] 🎯 Cliff: Learning Process Rewards from the First Mistake（Cliff：从第一个错误中学习过程奖励）[10:05] 🎯 Aspire: Can Models Self-Evolve from Vague Goals?（Aspire：模型能否从模糊目标中自我进化？）[10:46] 🎭 Influence-Directed Distillation: Solving the Diversity Bottleneck in Sampled-Token On-Policy Distillation（影响导向蒸馏：解决采样词元在线策略蒸馏中的多样性瓶颈）[11:42] 🎮 S3Gym: Can LLMs Turn Self-Testing and Self-Judging into Self-Improvement?（S3Gym：大语言模型能否将自我测试与自我评判转化为自我改进？）[12:36] 👀 A Glance Is All You Need: Single-Pass Fine-Grained Image Captioning with SimLoss（一瞥足矣：基于SimLoss的单遍细粒度图像描述）[13:35] 🎙 VibeVoice-ASR-Streaming Technical Report（VibeVoice-ASR-Streaming技术报告）【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

【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] 🎓 StudentSim: Training LLM-based Student Simulators（StudentSim：训练基于大语言模型的学生模拟器）[01:34] 🚗 Qwen-Drive-1.0: An Initial Step towards a Vision-Language Foundation Model for Autonomous Driving（Qwen-Drive-1.0：迈向自动驾驶视觉语言基础模型的第一步）[02:46] 🔁 SMELT: Scaling Laws for Compute-Matched MoE Looped Transformers（SMELT：计算匹配的MoE循环Transformer的缩放定律）[03:44] 🤖 UI-Venus-2 Technical Report（UI-Venus-2技术报告）[04:41] 🤖 ZimaBlue: Evolving Generalizable World Action Models through Scalable Video Pre-training（ZimaBlue：通过可扩展视频预训练演化可泛化的世界动作模型）[05:40] 🌍 H3-World: Turning Language Understanding into World Control（H3-World：将语言理解转化为世界控制）[06:41] 🔍 Hi-Q: Hierarchical Evidence-guided Query Refinement for Multi-Hop Question Answering（Hi-Q：层次化证据引导的多跳问答查询细化）[07:36] 🚁 Evaluating Multimodal LLMs as Generalist Vision-Language-Action Agents for Drone Control: Commanding, Approaching, Tracking and Searching（评估多模态大语言模型作为无人机控制的通用视觉-语言-行动智能体：指挥、接近、跟踪与搜索）[08:33] 🔗 Uncovering Understanding-Generation Synergy in Native Unified Multimodal Models: From Representation, Task to System（揭示原生统一多模态模型中理解与生成的协同效应：从表征、任务到系统）[09:27] 🛡 Safin-1: Safety from Within through Memory-Native State Evolution（Safin-1：通过记忆原生状态演化实现由内而外的安全）[10:24] 🏢 From Production Traffic to Post-Training: Building a Self-Hosted LLM That Covers the Corporate Request Mix（从生产流量到后训练：构建覆盖企业请求组合的自托管大语言模型）[11:15] 🩺 DiagEvo: Diagnosis-Guided Self-Evolution via Hierarchical Error Memory（DiagEvo：通过分层错误记忆进行诊断引导的自我进化）[12:07] 🧠 EM^2Mem: Event-Centric Multimodal Memory for Large Language Models（EM²Mem：面向大语言模型的事件中心多模态记忆）[13:09] 🤖 Harness-of-Harness: Multi-Day Autonomous Software Development with Continual Improvement（Harness-of-Harness：多日持续改进的自主软件开发）[14:07] 🛡 Control-Data Flow Separation: Stable Prompt Optimization in Multi-Agent LLMs（控制流与数据流分离：多智能体大语言模型中的稳定提示优化）【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

【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：[00:30] 🔄 Does On-Policy Distillation Really Distill? From Noisy Teacher to Self-Improvement（在线策略蒸馏真的在蒸馏吗？从噪声教师到自我改进）[01:24] 🎬 DreamX-Creator: Democratizing Native Audio-Video Generation at 2K Resolution（DreamX-Creator：在2K分辨率下实现原生音视频生成的民主化）[02:13] 🧠 GenFirst: Generation Before Reconstruction for Stable End-to-End Latent Generative Modeling（GenFirst：先生成后重建的稳定端到端潜在生成建模）[03:06] 🧩 Lucida: Parse, Generate, and Place for Composable Real-to-Sim Scene Modeling（Lucida：面向可组合真实到仿真场景建模的解析、生成与放置）[03:57] ⚖ Normalized Low-Rank Adaptation（归一化低秩自适应）[04:54] 🎯 PaperGym: Rubric-Centered Evolution for Research-Plan Generation（PaperGym：以评分标准为中心的研究计划生成进化）[05:53] ⚡ On the Design of Qwen3.8-Next Architecture: Evaluation, Efficiency, and Training Stability（论Qwen3.8-Next架构的设计：评估、效率与训练稳定性）[06:51] 🎓 CogEvol: Towards Efficient and Reliable Learning Environment Generation（CogEvol：迈向高效可靠的学习环境生成）[07:50] 🧭 LightNav-0: Eliciting VLM Spatial Intelligence for Generalist Embodied Navigation（LightNav-0：激发视觉语言模型空间智能，实现通用具身导航）[08:43] 🧩 SHAPE of Chain-of-Thought in Math Reasoning（数学推理中思维链的SHAPE框架）[09:48] 🧠 Scaling Large Reasoning Models beyond Human Supervision: A Path toward Superintelligence（超越人类监督的大规模推理模型扩展：通往超级智能之路）[10:51] 🧩 Super Library Agent: Joint Generation and Maintenance of Multiple Applications Beyond the Single Codebase（超级库智能体：超越单一代码库的多应用联合生成与维护）[11:48] 🤖 Evaluating the Hidden Costs of Personalization in Large Language Models（评估大型语言模型中个性化的隐藏成本）[12:42] 📋 Learning to Evaluate Before Improving: Automatic Rubric Induction for Automatic Research Agents（先评估后改进：面向自动科研智能体的自动评分标准归纳）[13:32] 🎭 Lies We Can See: Joint Verbal and Non-Verbal Deception by VLM Agents in Embodied Social Interactions（看得见的谎言：具身社会交互中VLM智能体的言语与非言语联合欺骗）【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

【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] 🔁 LoopArena: Benchmarking Models as Runtime Controllers for Loop Engineering（LoopArena：将模型作为循环工程的运行时控制器进行基准测试）[01:27] 💎 DART-SD: Diamond-topology Aware Retrieval and Tuning for Self-Distillation of Multi-Turn Tool-Calling Agents（DART-SD：多轮工具调用代理的菱形拓扑感知检索与自蒸馏调优）[02:25] 🛠 Agentic Artifact Creation: Systems, Evaluation, Principles, and Opportunities（智能体工件创建：系统、评估、原则与机遇）[03:25] 🤖 Beyond Data Scaling: Representation-Centric Continued Pre-training for Vision-Language-Action Models（超越数据规模：面向视觉-语言-动作模型的以表征为中心的持续预训练）[04:42] 🌍 Code as Worlds: Agentic Discovery of Executable World Representations for Physical Reasoning（代码即世界：面向物理推理的可执行世界表示的智能体发现）[05:45] 🔄 J-Zero: Unified Challenger--Solver--Judge Co-Evolution from Zero Data（J-Zero：零数据下挑战者—求解者—评判者的统一协同进化）[06:37] 🎥 Revisiting Local Context for Long-Horizon Streaming 3D Reconstruction（重新审视面向长时程流式三维重建的局部上下文）[07:35] 🧠 ContextPilot: Teaching Agents for Proactive Context Management via Fine-grained RL（ContextPilot：通过细粒度强化学习训练智能体进行主动上下文管理）[08:33] 🧠 LayerRecall: A State-Conditioned Memory Router for Long-Horizon Consistency in Video Generation（LayerRecall：用于视频生成长时程一致性的状态条件记忆路由器）[09:38] 💰 Puro-2B: Poor Lab's Qwen2-1.5B Trained on RTX 5090 within $5090（Puro-2B：穷实验室在RTX 5090上以不到5090美元训练出的Qwen2-1.5B）[10:27] 🐘 Blind Men and the Elephant: Probing the Epistemic Myopia of LLMs under Long-Tail Divergent Knowledge（盲人摸象：探究长尾分歧知识下大语言模型的认识论短视）[11:18] 🛡 StepGuard: Learning Step-Level Guardrails with Scalable Supervision and Safety-Utility Balancing（StepGuard：通过可扩展监督与安全效用平衡学习步骤级护栏）[12:12] 🎨 Paint What You See: Benchmarking Dexterous Visual Tool Use in Multimodal Agents（画你所见：多模态智能体中灵巧视觉工具使用的基准测试）[13:08] ⚡ Locate Anything in Videos: Rethinking Efficient Generative Spatio-Temporal Video Grounding（在视频中定位一切：重新思考高效生成式时空视频定位）[14:04] 🧠 PonderPounce: A Pretrained MLLM as an Episode Context Engine for Robot Control（PonderPounce：预训练多模态大语言模型作为机器人控制的回合上下文引擎）【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

【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：[00:42] TOP1(🔥253) | 🧠 VBVR-Pro: A Scalable and Verifiable Suite for Native Visual Reasoning（VBVR-Pro：可扩展且可验证的原生视觉推理套件）[03:58] TOP2(🔥201) | 🤖 Apodex 1.1: Scaling Agentic Intelligence for Complex Work（Apodex 1.1：扩展智能体智能以应对复杂工作）[07:26] TOP3(🔥170) | 🎬 VGI-Bench: Probing Visual Intelligence in Video Generation Models（VGI-Bench：探究视频生成模型的视觉智能）[10:06] TOP4(🔥167) | 🧠 VoiceMem: Streaming Dual-Brain Memory for Real-Time Interaction（VoiceMem：面向实时交互的流式双脑记忆）[13:16] TOP5(🔥140) | 🧪 FrontierChallenge: Evaluating Scientific Workflow Completion（前沿挑战：评估科学工作流完成度）【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

【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：[00:30] 🎲 PAWBench: How Far Are We from Probabilistically Aligned World Modeling?（PAWBench：我们距离概率对齐的世界建模还有多远？）[01:09] 🤖 What Makes Good Agentic Data? An ACE Lens on Data Generation for LLM Agents（什么造就了好的智能体数据？面向LLM智能体的数据生成的ACE视角）[02:01] 🎯 TTPO: Test-Time Policy Optimization（TTPO：测试时策略优化）[02:49] 🤖 Training Agents to Evolve with Their Harness: TaoLive Digital Avatar Agent Technical Report（训练智能体与其运行框架协同进化：TaoLive数字人智能体技术报告）[03:41] 🏙 UrbanGround: From Local Perception to Spatial Agency in a Real-Scale City（UrbanGround：从局部感知到真实尺度城市中的空间能动性）[04:34] 🎮 GameWAM: A World Action Model for Video Games（GameWAM：视频游戏的世界动作模型）[05:24] 🔄 PILOT in the Loop: Live Self-Improvement for Long-Horizon Agents（PILOT在环：面向长时程智能体的实时自我改进）[06:03] 🤖 Zero-WAM: In-Context World-Action Modeling from Human Videos for Open-Ended Task Generalization（Zero-WAM：基于人类视频的上下文世界-动作建模实现开放式任务泛化）[06:55] ⚗ Self-OPD: On-Policy Distillation for Flow Matching Models without Teacher（Self-OPD：无需教师模型的流匹配模型在线策略蒸馏）[07:51] 🧬 Understanding Evolution Strategies for LLM Reasoning: Broader Reasoning Coverage than GRPO（理解面向LLM推理的进化策略：比GRPO更广泛的推理覆盖）[08:44] 🎮 Magpie: Real-Time World Renderer for Interactive Games（Magpie：用于交互式游戏的实时世界渲染器）[09:41] 🧩 Procedura: Agentic 3D Modeling with Procedural Control（Procedura：程序化控制下的智能体3D建模）[10:39] 🎬 Thinking on Shots: Consistent Multi-Shot Video Editing with Agentic Reasoning（思考镜头：基于智能体推理的一致性多镜头视频编辑）[11:40] 🧠 WikiSkill: Compiling Agent Experience into Persistent Knowledge for Skill Evolution（WikiSkill：将智能体经验编译为持久知识以促进技能进化）[12:46] 🔗 CaSKG: Counterfactual-Causal Skill Graphs for Scalable Agent Skill Retrieval（CaSKG：面向可扩展智能体技能检索的反事实因果技能图）【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

【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：[00:31] 🧠 VoiceMem: Streaming Dual-Brain Memory for Real-Time Interaction（VoiceMem：面向实时交互的流式双脑记忆）[01:30] 🧪 FrontierChallenge: Evaluating Scientific Workflow Completion（前沿挑战：评估科学工作流完成度）[02:29] 🎬 VGI-Bench: Probing Visual Intelligence in Video Generation Models（VGI-Bench：探究视频生成模型的视觉智能）[03:32] 🚀 WarpSAC: Towards the Pinnacle of Scalable Off-policy RL by Rethinking Exploration and Exploitation（WarpSAC：重思探索与利用，迈向可扩展离策略强化学习的巅峰）[04:27] ⚙ JIT-Agent: Scaling Harness Intelligence via Just-in-Time Harness Evolution（JIT-Agent：通过即时框架演化实现框架智能的规模化扩展）[05:20] 🧠 VBVR-Pro: A Scalable and Verifiable Suite for Native Visual Reasoning（VBVR-Pro：可扩展且可验证的原生视觉推理套件）[06:15] 🎛 D$^3$-MOPD: Adaptive Dynamic Domain ScheDuling for Efficient Multi-Teacher Distillation（D³-MOPD：面向高效多教师蒸馏的自适应动态领域调度）[07:22] 🧠 Is Next-Chunk Reasoning RL Really Better than SFT? Revisiting Training Strategies under no-CoT Data（下一块推理强化学习真的比SFT更好吗？重新审视无CoT数据下的训练策略）[08:26] 🎯 Agent-G$^2$: Gaussian Guidance for Agentic Reinforcement Learning（Agent-G²：面向智能体强化学习的高斯引导）[09:13] 🎬 Long-Horizon Audio-Visual Generation for Persistent Stories and Interactive Worlds（面向持久故事与交互世界的长时程音视频生成）[10:19] ⚖ Open-MOPD: Diagnosing and Fixing Capability Imbalance in Multi-Teacher On-Policy Distillation（Open-MOPD：多教师同策略蒸馏中能力失衡的诊断与修复）[11:26] 🤖 StreamPI: Streaming Multimodal Temporal Modeling for Vision-Language-Action Models（StreamPI：面向视觉-语言-动作模型的流式多模态时序建模）[12:29] 🎬 Video-IFBench: Evaluating Instruction Following of Multimodal LLMs in Video Understanding Scenarios（视频-IFBench：评估多模态大语言模型在视频理解场景中的指令跟随能力）[13:30] 🧠 Code World Model: Coding Agent as World Brain（代码世界模型：代码智能体作为世界大脑）[14:25] 🎯 V-Rubrics: Visual Faithfulness via Rubric-Based Reinforcement Learning（V-Rubrics：基于评分标准的强化学习实现视觉忠实性）【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

【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：[00:32] 🎯 Annotations as Rollouts: Efficient and Scalable Reinforcement Learning for Video MLLMs（将标注视为轨迹：高效可扩展的视频多模态大语言模型强化学习）[01:25] 🧲 WeMM-Embedding: WeChat Multi-Modal Embedding Technical Report（WeMM-Embedding：微信多模态嵌入技术报告）[02:17] 🔧 AutoSaddler: Automatic Harness Optimization with Durable Updates from Agent Execution Traces（AutoSaddler：基于智能体执行轨迹的自动框架优化与持久更新）[03:17] 🎯 On-Policy Self-Distillation in Diffusion Models（扩散模型中的同策略自蒸馏）[04:02] 🛡 CyberFactory: Scaling Cyber Security Capabilities with Instances from the Wild（CyberFactory：利用现实世界实例规模化网络安全能力）[05:00] 🧠 Recursive Experiential-Working Memory Evolution for Long-Horizon Agent Harnesses（面向长时程智能体框架的递归经验-工作记忆演化）[05:53] 🎯 Best Practice Critic Optimization（最佳实践评论家优化）[06:45] 🎯 On-policy Distillation with Verifiable Reward（基于可验证奖励的在线策略蒸馏）[07:42] 🕶 From Seeing to Acting: Smart Glasses as First-Person Intelligence Platforms（从看见到行动：智能眼镜作为第一人称智能平台）[08:41] 🎮 Game2World Engine: Unlocking In-the-Wild Gameplay Videos for World Model Training（Game2World引擎：解锁真实场景游戏视频以训练世界模型）[09:34] 📏 Length-Adaptive Decoding for Masked Diffusion Machine Translation（掩码扩散机器翻译的长度自适应解码）[10:27] 🎬 LAION-BVD: A 10-Million-Hour Open Video Dataset for Multimodal Pre-training（LAION-BVD：面向多模态预训练的千万小时开放视频数据集）[11:21] 🔁 Meta$^n$: Recursive Self-Improvement through Emergent Depth（Meta^n：通过涌现深度实现递归自我改进）[12:13] 🤖 CAFE: Self-Improving Search Agents Need Co-Evolving Feedback（CAFE：自我改进的搜索智能体需要协同演化的反馈）[13:14] 🔮 Latent Action as Intention Enables Efficient Future Imagination for World Action Models（以潜在动作作为意图，实现世界动作模型的高效未来想象）【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

【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：[00:30] 🤖 Apodex 1.1: Scaling Agentic Intelligence for Complex Work（Apodex 1.1：扩展智能体智能以应对复杂工作）[01:36] 🎮 EchoWM: Open and Enterable Omnimodal World Models（EchoWM：开放且可进入的全模态世界模型）[02:40] 🛒 TLive-Omni: An Omni-Modal Understanding Model for E-Commerce Live Streaming（TLive-Omni：面向电商直播的全模态理解模型）[03:22] 🎨 Unlocking the Potential of Image Editing via Concept Scaling and Dense Supervision（通过概念缩放与密集监督解锁图像编辑的潜力）[04:20] 📱 MobilePA-Bench: Benchmarking Mobile Planner Agents on Complex Real-World Tasks（MobilePA-Bench：面向复杂真实世界任务的移动规划智能体基准评测）[05:16] 🤖 Prime Agent: A Self-Improving RLM Harness（Prime Agent：一种自我改进的递归语言模型工具框架）[06:22] 🧊 Block3D: Efficient Text-to-3D Generation via Block-Wise Diffusion（Block3D：通过块状扩散的高效文本到三维生成）[07:09] 🚗 RISE: Adaptive Imagination for World Action Models（RISE：面向世界行动模型的自适应想象）[08:19] 📈 Towards a Densing Law for User Representation Learning at Billion-Scale Capacity（面向十亿级容量用户表示学习的稠密化定律）[09:26] ⚖ ARC: Fair Relative Advantage Comparison in Open-Ended Real-World Interaction（ARC：开放式真实世界交互中的公平相对优势比较）[10:27] 🧠 ReWorld: An Interactive World Model with Long-Horizon Memory（ReWorld：一个具有长时记忆的交互式世界模型）[11:29] ⚖ Beyond the Stability-Exploration Dilemma: Environmental Regularization for LLM Policy Optimization（超越稳定性-探索困境：面向大语言模型策略优化的环境正则化）[12:23] 🎮 GameXpert-Bench: How Far Are Coding Agents from Expert Game Development?（GameXpert-Bench：编码智能体距离专家级游戏开发还有多远？）[13:17] 🧪 One Success Isn't Reliability: Thinkingbox, a Sandbox and Benchmark for Agents in Stateful Business Workflows（一次成功不等于可靠：Thinkingbox——面向有状态业务工作流的智能体沙盒与基准测试）[14:26] 🎯 Task-CoEvolve: Efficient Harness Optimization via Adaptive Validation Task Selection（Task-CoEvolve：通过自适应验证任务选择实现高效的工具链优化）【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

【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：[00:32] 🚀 Let's Scale Step by Step: Compute-Efficient Hyperparameter Transfer for Large-Scale Mixture-of-Experts（让我们一步一步扩展规模：面向大规模混合专家模型的高效超参数迁移）[01:24] 🕸 Graph Engineering in the Era of LLM Agents: From Individual Intelligence to System Intelligence（大语言模型智能体时代的图工程：从个体智能到系统智能）[02:21] 🤖 OmniAssistBench: Assistant-style Interaction Benchmark for Omni-LLMs（OmniAssistBench：全模态大语言模型的助手式交互基准）[03:17] ⚡ ParaTempo: Efficient Parallel Reasoning via Temporal Confidence（ParaTempo：基于时间置信度的高效并行推理）[04:00] ♾ InfinityEdit: Infinite Video Editing with a Lightweight Edit-Ignition Adapter（InfinityEdit：基于轻量级编辑触发适配器的无限视频编辑）[04:53] 🪙 Every Coin Has Two Sides: On the Dual Nature of Generalization in On-Policy Distillation of Large Language Models（每一枚硬币都有两面：论大型语言模型同策略蒸馏中泛化的双重性）[05:37] 🧩 EviRank: Structured Relevance Evidence for Multimodal Image Re-ranking（EviRank：面向多模态图像重排序的结构化相关性证据）[06:46] 🧠 Beyond Correctness: Benchmarking and Aligning Response Behaviors in Hybrid-Thinking MLLMs（超越正确性：混合思考多模态大语言模型的响应行为基准测试与对齐）[07:45] 🎨 UniSpace: Unified Visual Representation and Scalable Multimodal Modeling（UniSpace：统一视觉表示与可扩展多模态建模）[08:39] 🛒 Towards Faithful Simulation of Human Shopping Behavior（面向人类购物行为的忠实模拟）[09:35] ⚙ AgentMercury: Your Agent Can Synthesize Verifiable Environments for Business Scenarios at scale（AgentMercury：你的智能体能够大规模合成可验证的商业场景环境）[10:41] ⚡ Daedalus-150M: A Convolution-Attention Hybrid Designed for CPU Inference（Daedalus-150M：为CPU推理设计的卷积-注意力混合模型）[11:35] 🛡 CLEAR: Continuous Latent Adapter Routing for Utility-Preserving LLM Safety Alignment（CLEAR：面向保持实用性的大语言模型安全对齐的连续潜在适配器路由）[12:42] 📱 Llama-Mobile: Efficient 2.7-Bit Quantization of VLMs（Llama-Mobile：高效2.7比特视觉语言模型量化）[13:32] 🍳 FlavourBench: Ranking Frontier Language Models with Executable Culinary Ground Truth（FlavourBench：用可执行的烹饪真值对前沿语言模型进行排名）【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

【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：[00:43] TOP1(🔥435) | ⚙ StateM: Reaching 95.3% Raw Accuracy, or a \$15 Frontier Run, on Terminal-Bench 2.1 via Harness Scaling（StateM：通过 Harness 扩展在 Terminal-Bench 2.1 上达到 95.3% 原始准确率，或一次 15 美元的前沿运行）[03:30] TOP2(🔥277) | 🛡 Can We Defend Against AI-Generated Video Attacks on Real-World Crisis Events? A Systematic Evaluation of Detectors, Generators and Social Dissemination（我们能防御针对现实世界危机事件的AI生成视频攻击吗？对检测器、生成器与社会传播的系统评估）[06:28] TOP3(🔥245) | 🌍 EnvHarness: Awakening Static Worlds for Agent Learning（EnvHarness：为智能体学习唤醒静态世界）[09:45] TOP4(🔥167) | 👁 Self-Supervised Visual On-Policy Distillation（自监督视觉同策略蒸馏）[12:21] TOP5(🔥162) | 🧩 Demystifying Agent Skills: Why They Work-Until They Don't（揭秘智能体技能：它们为何有效——直到失效）【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

【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：[00:31] 🌍 EnvHarness: Awakening Static Worlds for Agent Learning（EnvHarness：为智能体学习唤醒静态世界）[01:28] 🖥 FACET: Preserving Source Intent and Executable State in Terminal Task Synthesis（FACET：在终端任务合成中保留源意图与可执行状态）[02:29] 🧪 SWE-bench Science: Can Coding Agents Resolve Engineering Tasks in Science?（SWE-bench Science：编码智能体能否解决科学领域的工程任务？）[03:15] 🕺 4DAnyone: Create Anyone in 4D from a Casual Monocular Video（4DAnyone：从随意单目视频创建任意人物的4D形象）[04:11] 👥 WithEveryone: Unified Planning and Identity Grounding for Group Image Generation（WithEveryone：面向群像生成的统一规划与身份锚定）[05:01] 🧠 MemTrapBench: Benchmarking Cognitive Traps in LLM Memory Use（MemTrapBench：大语言模型记忆使用中的认知陷阱基准测试）[05:52] 🔄 SkillEvo: Self-Renewing Evolution Gradients from Multi-Turn Interaction Feedback（SkillEvo：来自多轮交互反馈的自我更新进化梯度）[06:50] 🎮 ForgeWM: Progressive Causal Training for Few-Step Action-Conditioned Video World Models（ForgeWM：面向少步动作条件视频世界模型的渐进式因果训练）[07:48] 🧩 Repo0: Design-Driven Zero-to-All Code Generation（Repo0：设计驱动的从零到全代码生成）[08:46] ⚡ FlashPrefill V2: Block-Sparse Prefill Attention for Long-Context LLM Serving（FlashPrefill V2：面向长上下文大语言模型服务的块稀疏预填充注意力）[09:41] 🧠 Inject, Align, Recover: Staged Post-Training for Retrieval-Free Document Knowledge Internalization（注入、对齐、恢复：面向无检索文档知识内化的分阶段后训练）[10:42] 🤖 EXIMO: VLM Guided Exploration of VLA Policies（EXIMO：视觉语言模型引导的VLA策略探索）[11:36] 🧠 Thinking in a Low-Resource Language: What SFT Builds, What RL Fixes, What Accuracy Cannot See（用低资源语言思考：SFT构建了什么，RL修复了什么，准确率无法看到什么）[12:18] 🎯 Towards Quantifying Benchmark Optimization in ASR Models（面向ASR模型中基准优化的量化研究）[13:20] 🛡 PolicyGuide: From Guarding One Action to Guiding the Whole Workflow for Policy-Compliant LLM Agents（PolicyGuide：从守护单一动作到引导策略合规型LLM智能体的整个工作流）【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

【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：[00:34] 🤖 Zetta $ζ$: An Efficient Closed-Loop Embodied Harness for Self-Evolving Physical Intelligence（Zetta ζ：面向自进化物理智能的高效闭环具身智能体框架）[01:29] ✅ SemaPLC: A Project-Grounded, Verification-Gated Agent Harness for PLC Code Generation（SemaPLC：一种基于项目、以验证为门控的PLC代码生成智能体框架）[02:31] 🎯 SemComp-Bench: Benchmarking Semantic Task Completion in Video Generation（SemComp-Bench：视频生成中语义任务完成的基准测试）[03:28] 🔬 OmniScientist: An Omni-Modal Omni-Discipline AI Scientist（全能科学家：一个全模态、全学科的人工智能科学家）[04:26] 🧠 Co-RL: Unsupervised Reasoning Emerges from Diverse Cohort in Multi-agent RL（Co-RL：多智能体强化学习中多样化群体催生无监督推理）[05:13] 🎮 SPADE: Self-Play in Adaptive Synthetic Executable Environments（SPADE：自适应合成可执行环境中的自博弈）[06:08] 🧪 Training Chemical Plausibility-Aware Large Language Models for Single-Step Retrosynthesis（训练面向单步逆合成的化学合理性感知大语言模型）[07:04] 🎯 Decision-Metric Alignment in Latent World Models: Diagnostics and Action-Conditioned Objectives for MPC Planning（潜在世界模型中的决策度量对齐：诊断与面向MPC规划的动作条件目标）[07:58] 🧬 Training Leaves Traces: Centered Residual Signatures for Language Model Lineage Verification（训练留下痕迹：面向语言模型谱系验证的中心化残差签名）[08:50] 🔁 Looped Language Models Improve Compositional Tool Calling（循环语言模型提升组合式工具调用能力）[09:36] 🖐 SoftVTBench: A Deformation-Aware Visuo-Tactile Dataset and Benchmark for Deformable-Object Manipulation（SoftVTBench：面向可变形物体操作的变形感知视触觉数据集与基准）[10:33] ⚽ FM-Bench: A Benchmark for Long-Horizon Management with Competing Agents（FM-Bench：面向竞争智能体的长时程管理基准）[11:25] ✍ Scaling Creative Writing Beyond Story-Centric Data with Attribute-Guided Genre Expansion（借助属性引导的体裁扩展，将创意写作扩展到故事中心数据之外）[12:15] 🔥 The More Popular, The Harder to Forget: Adaptive Popularity for LLM Unlearning（越热门越难遗忘：大语言模型遗忘的自适应流行度方法）[13:01] 🔍 Temporal Multi-Signal Fusion for Token-Level Hallucination Detection（面向Token级幻觉检测的时序多信号融合）【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

【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：[00:34] 🧬 Agentic ESOpt: Fine-Tuning Long-Horizon LLM Agents with Minimal GPU Requirements（Agentic ESOpt：以极低GPU需求微调长程LLM智能体）[01:46] 🧩 Demystifying Agent Skills: Why They Work-Until They Don't（揭秘智能体技能：它们为何有效——直到失效）[02:35] 🔬 ASI-Bench: At the Dawn of Artificial Superintelligence（ASI-Bench：人工超级智能的黎明）[03:26] 💻 FreeToken: Efficient Edge-Native MoE Serving with Bandwidth-Adaptive Execution（FreeToken：高效的边缘原生MoE服务与带宽自适应执行）[04:20] 🧭 Embodied-Navigator: Point, Think, Memorize, and Align for Efficient Navigation（具身导航器：指向、思考、记忆与对齐实现高效导航）[05:22] 🎬 AVA-Encoder: Towards Agent-Native Video Representation Learning（AVA-编码器：迈向智能体原生的视频表示学习）[06:12] 🖼 EDITBRIDGE: Towards Faithful and Efficient Ultra-High-Resolution Image Editing（EDITBRIDGE：迈向忠实且高效的超高分辨率图像编辑）[07:13] ⚡ Agent Lightning v1.0: Towards Harnessed Agentic RL（Agent Lightning v1.0：迈向框架化的智能体强化学习）[08:12] 🎥 V-RAE: Rethinking Video Latent Spaces for Generation（V-RAE：重新思考用于生成的视频潜空间）[09:17] 🎬 CoinVE-200K: A Large-Scale High-Quality Dataset for Compositional Instruction-Guided Video Editing（CoinVE-200K：面向组合式指令引导视频编辑的大规模高质量数据集）[10:17] 🛡 DiSCO: Defending text-to-image generation through distribution-guided contrastive prompt optimization（DiSCO：通过分布引导的对比提示优化防御文本到图像生成）[11:22] 🧠 Harness the Memory: A Holistic Evaluation of Memory Substrates in Memory Agents（驾驭记忆：对记忆智能体中记忆底层介质的整体评估）[12:32] 🎨 From Corpora to Co-Evolving Capabilities: Capability-Centric Data Design for Generalist Image Generation（从语料到协同演进的能力：以能力为中心的通用图像生成数据设计）[13:31] 📊 StartupBench: Benchmarking General-Purpose Agents on Market-Validated End-to-End Workflows（StartupBench：对通用智能体在市场验证的端到端工作流上的基准测评）[14:34] ⚡ Energy-Guided Flow Matching（能量引导的流匹配）【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

【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：[00:32] 🕵 HarnessEval-W: Agentifying the Evaluation of Visual Worlds（HarnessEval-W：使视觉世界的评估智能体化）[01:32] 🌍 VibeWorlding: Can Multimodal Agents Construct 3D Open Worlds End-to-End?（VibeWorlding：多模态智能体能否端到端构建3D开放世界？）[02:10] ⚡ MOSS-VL Technical Report（MOSS-VL 技术报告）[03:05] 🤖 ClawGym II: Exploring Black-Box RL on Agent Harness（ClawGym II：在智能体框架上探索黑盒强化学习）[03:56] 🎯 Learn What's Left, Not What's Mastered: Saturation Aware Advantage Reweighting for Multi-Reward Policy Optimization（学习尚未掌握的，而非已经精通的：面向多奖励策略优化的饱和感知优势重加权）[04:51] 🤖 UI-Mate: Advancing Open-Weight Foundation GUI Agents with In-Context Demonstrations（UI-Mate：利用上下文演示推进开放权重的基础图形用户界面智能体）[05:46] 🔬 Large Discovery Models: Empirically-grounded Model-Based Open-Ended Search（大型发现模型：基于经验建模的开放式搜索）[06:37] 🎨 An Empirical Study of Training Pixel-Space Text-to-Image Diffusion Models（训练像素空间文本到图像扩散模型的实证研究）[07:31] 🤖 Agentic Transaction: Towards ACID-Compliant Agent Systems（智能体事务：迈向ACID合规的智能体系统）[08:29] 🔬 How Do Agents Fail on AutoResearch: End-to-End Diagnostic Evaluation on 100 Real-World Frontier Research Tasks（智能体如何在自动研究中失败：基于100个真实前沿研究任务的端到端诊断评估）[09:24] ⚡ GenRouter: Unified Workflow Routing for Agentic Image Generation（GenRouter：用于智能体图像生成的统一工作流路由）[10:29] 🧩 MegaParts: Scaling Part-Aware 3D Object Generation to 300 Parts via Token-Efficient Autoregressive Modeling（MegaParts：通过词元高效的自回归建模将部件感知的3D物体生成扩展到300个部件）[11:22] 🧠 Understanding Cognition-Induced Risks in Agentic AI Systems（理解智能体AI系统中认知引发的风险）[12:21] 🔗 Advancing Open and Reproducible Relational Learning: RelArena-$α$, TabPFN-Rel and RPI（推进开放可复现的关系学习：RelArena-α、TabPFN-Rel与RPI）[13:24] 🛡 Ventor-QTest: Threat-Model-Driven Verification of Vendor-Hosted LLM APIs（Ventor-QTest：威胁模型驱动的供应商托管LLM API验证）【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

【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：[00:30] 🛡 Can We Defend Against AI-Generated Video Attacks on Real-World Crisis Events? A Systematic Evaluation of Detectors, Generators and Social Dissemination（我们能防御针对现实世界危机事件的AI生成视频攻击吗？对检测器、生成器与社会传播的系统评估）[01:27] 👁 Self-Supervised Visual On-Policy Distillation（自监督视觉同策略蒸馏）[02:21] 🤖 Beyond Final Scores: A Systematic Evaluation of Agents for Long-Horizon AI Research and Development（超越最终得分：对长周期AI研究与开发智能体的系统评估）[03:15] 🧠 Intern-S2-Mobius: Foundation Model with Decoupled Knowledge and Reasoning（Intern-S2-Mobius：知识与推理解耦的基础模型）[04:09] 🎮 Marionette: Predicting World States, Rendering Geometry, Painting Appearance（Marionette：预测世界状态，渲染几何，绘制外观）[05:00] 🧠 SimpleOPD: Simple Tokenizer-Agnostic On-Policy Distillation for Long-Context Reasoning（SimpleOPD：面向长上下文推理的简单分词器无关在线策略蒸馏）[06:03] 🧠 MobileMem: Learning from a Year of Mobile Experiences（移动记忆：从一年的移动体验中学习）[06:54] 🤖 DFM Mimir v1: An Open HRM Delivering Frontier Performance at 1B Parameters Using Only Permissible Post-Training Data（DFM Mimir v1：仅使用合规后训练数据、以1B参数实现前沿性能的开放HRM）[07:50] 🤸 HumanTracker: Towards Comprehensive and Human-Aligned Motion Tracking Benchmark（HumanTracker：迈向全面且与人类感知一致的运动追踪基准）[08:50] 🎨 CPI-Bench: A Comprehensive,Practical and Intelligent Benchmark for Real-World Image Editing（CPI-Bench：一个面向真实世界图像编辑的全面、实用且智能的基准）[09:48] 🧠 Latent On-Policy Self-Distillation（潜在同策略自蒸馏）[10:53] 🔍 Claim-Level Reliability Assessment for Efficient Test-Time Reasoning（面向高效测试时推理的声明级可靠性评估）[11:43] 🤖 PRM-as-a-Judge 1.5: A Toolkit for Robot Process Assessment（PRM-as-a-Judge 1.5：机器人过程评估工具包）[12:42] 📉 Forecast Collapse in Time-Series Foundation Models（时间序列基础模型中的预测崩溃）[13:41] 🤔 Second Thought: Reasoning in Parallel as LLM Agents Act and Observe（第二思考：LLM智能体在行动与观察时并行推理）【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

【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：[00:46] TOP1(🔥639) | 🧠 BDH-CQ: In-Context Learning with Recurrent Latent Reasoning（BDH-CQ：基于循环潜在推理的上下文学习）[03:27] TOP2(🔥332) | 🔄 Macaron-V1: Towards Open Continual Learning with Self-Improvement and Mixture-of-LoRA（Macaron-V1：迈向具备自我改进和LoRA混合的开放持续学习）[06:39] TOP3(🔥278) | 📄 Spark-to-Paper: End-to-End Research Paper Generation as a Composable Skill（从火花到论文：作为可组合技能的端到端研究论文生成）[10:15] TOP4(🔥255) | 🧬 OpenART: Scaling Agent Red Teaming via Open-Ended Environment Evolution（OpenART：通过开放式环境演化扩展智能体红队测试）[13:01] TOP5(🔥208) | 🧠 On-Policy Self-Distillation without Any Supervision（无需任何监督的在线策略自蒸馏）【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

【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：[00:32] 🤖 DreamX-Phi 1.0: Action-Conditioned Video World Model for Robotic Manipulation（DreamX-Phi 1.0：面向机器人操作的动作条件视频世界模型）[01:24] 🌍 Alaya-EVOKE: From Linear-Scaling Supervision to Endless World（Alaya-EVOKE：从线性扩展监督到无尽世界）[02:30] 🔀 LLMRouter: Unified Infrastructure for Developing, Evaluating, and Deploying LLM Routers（LLMRouter：开发、评估和部署LLM路由器的统一基础设施）[03:25] 🧬 DarwinX: Evolving Agent Harnesses Through Natural Selection（DarwinX：通过自然选择进化智能体框架）[04:26] 🔬 Intern-S2-Preview: Scientific Agentic Foundation Model（Intern-S2-Preview：科学智能体基础模型）[05:20] 🎮 PlayWorld: Benchmarking World Models with Agent Players over Long-Horizon Objectives（PlayWorld：使用智能体玩家在长程目标上对世界模型进行基准测试）[06:20] 🤖 AutoDesign: Meta-Harness Optimization for Long-Horizon Agentic Design（AutoDesign：面向长时程智能体设计的元框架优化）[07:23] 🧠 Spatial Memory Agent: Experience-Grounded Procedure Memory for Spatial Intelligence（空间记忆智能体：基于经验的程序记忆实现空间智能）[08:12] ⚡ Massive Activations in Hybrid Linear Attention Large Language Models: Pre-Attention Spikes and Inter-Spike Plateaus（混合线性注意力大语言模型中的大规模激活：注意力前尖峰与尖峰间平台）[09:03] 🎭 UniSwap: Streaming Audio-Visual Identity Swapping for Talking Videos（UniSwap：面向说话视频的流式音频-视觉身份交换）[10:12] ⚡ LiveAnimate: Stable Long-Form Streaming Human Animation in Real-Time（LiveAnimate：实时稳定长格式流式人体动画生成）[11:08] 🤖 How Can Rhetoric Reward-Hack AI Reviewers? Dissecting Rhetorical Sensitivity in AI-Based Peer Review（修辞何以能对AI审稿人进行奖励黑客？解析基于AI的同行评审中的修辞敏感性）[12:00] ✂ An AI4AI Framework for Visual Token Pruning（面向视觉Token剪枝的AI4AI框架）[12:56] 🤖 H2R-Bench: Benchmarking Human-to-Robot Manipulation Video Generation in World Models（H2R-Bench：在世界模型中评估人类到机器人的操作视频生成）[14:08] 🔄 Full-bandwidth transformer（全带宽Transformer）【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

【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：[00:26] 🧬 OpenART: Scaling Agent Red Teaming via Open-Ended Environment Evolution（OpenART：通过开放式环境演化扩展智能体红队测试）[01:35] 📄 Spark-to-Paper: End-to-End Research Paper Generation as a Composable Skill（从火花到论文：作为可组合技能的端到端研究论文生成）[02:43] 🧩 AI4AI at Test-Time: Strong-to-Weak Capability Transfer via Harnesses（测试时AI4AI：通过推理支架实现强到弱能力迁移）[03:37] 🔬 Mechanist: AI as a Scientific Instrument for Discovering the Mechanisms of Intelligence（Mechanist：将人工智能作为揭示智能机制的科学仪器）[04:33] 🎭 Can LLM Agents Stick to the Script? A Benchmark for Long-Horizon Consistency in Interactive Narratives（大语言模型智能体能否坚守剧本？交互式叙事中长程一致性的基准测试）[05:33] 🌍 StateFlow: Building, Evolving, and Accessing 3D World States for Previsualization（StateFlow：为预可视化构建、演化与访问3D世界状态）[06:28] 📐 Self-Geometry: GT-Free and Plug-and-Play Test-Time Adaptation for Geometrically Consistent 3D Vision Foundation Models（自几何：面向几何一致的3D视觉基础模型的无真值即插即用测试时自适应）[07:25] 🪞 From Synthesis to Removal: Physics-Grounded Reflection Simulation and Diffusion-Based Video Dereflection（从合成到去除：物理驱动的反射模拟与基于扩散模型的视频去反射）[08:22] 🛡 ToolHazard: Scaling Adversarial Environments for Security Evaluation and Alignment of LLM-based Agents（ToolHazard：扩展对抗性环境，用于基于大语言模型的智能体的安全评估与对齐）[09:15] 🔍 The Illusion of Visual Tool-Use: A Causal Audit of Thinking with Images（视觉工具使用的幻象：图像思维的因果审计）[10:16] 🤖 Self-Evolving Embodied Agents via Skill-Harness Evolution（基于技能与执行框架演化的自进化具身智能体）[11:19] 🧩 SkillZip: Contract-Preserving Graph Compression for Scalable Agent Skill Libraries（SkillZip：面向可扩展智能体技能库的合约保持图压缩）[12:14] 🛡 Agent Safety Should Be a Runtime Contract（智能体安全应当是一种运行时契约）[13:01] 🧬 Persistent Recursive Worlds Enable Autonomous Software Evolution（持久递归世界赋能自主软件演化）[13:46] 💡 MBA: Multimodal Benchmark and Agents for Real-World Business Ideation（MBA：面向真实世界商业构思的多模态基准与智能体）【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

【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：[00:35] 🤖 ComBodied Agents: a New Paradigm of Human-Centric Agentic AI（共体智能体：以人为中心的智能体人工智能新范式）[01:22] 🧬 Co-Evolution in Agentic Systems: Toward Self-Directed Evolution Beyond Human Design（智能体系统中的共同演化：迈向超越人类设计的自我导向演化）[02:22] 🌍 Beyond Pixels: From Video Priors to 4D Worlds（超越像素：从视频先验到4D世界）[03:12] 🧩 Articulated Object Reconstruction from Rest-State Observation（基于静止状态观测的铰接物体重建）[04:09] ⚔ AdvFD: Boosting Visual Generation via Adversarial Fr'echet Distance Loss（AdvFD：通过对抗性弗雷歇距离损失提升视觉生成）[05:05] 🧬 Mendel Gödel Machine: Recursive Self-Improving Coding Agents via Comparative Evolution（孟德尔·哥德尔机：通过比较进化实现递归自我改进的编码智能体）[06:00] 🎭 Ex-Omni-2D: Expressive Omni-Modal Dialogue Models with Native Visual Presence（Ex-Omni-2D：具备原生视觉临场感的表现性全模态对话模型）[06:48] 🌍 VibeLifeBench: Can Your Life Agent Be Proactive and Persistent in a Living World?（VibeLifeBench：你的生活智能体能否在动态世界中主动且持久地行动？）[07:46] 🚫 Decoding-Level Taboo: A Diagnostic Stress Test for LLM Robustness（解码级禁忌：大语言模型鲁棒性的诊断性压力测试）[08:33] 📦 SkillZip: Evaluation-Free Skill Compression for Self-Evolving Agents by Discovering Reusable Structure（SkillZip：通过发现可复用结构实现自我进化智能体的免评估技能压缩）[09:33] 📱 SPIEval: Evaluating Large Language Models as Mobile Assistants over Scattered Personal Information（SPIEval：评估大型语言模型作为移动助手处理分散个人信息的能力）[10:36] ✂ Not Worth Another Token: Marginal Value Estimation for Efficient Deep Research Agents（并不值得再投入一个词元：高效深度研究智能体的边际价值估计）[11:34] 🌐 Reference-Free Post-Training of Open Large Language Models for Multilingual Machine Translation（开放大语言模型用于多语言机器翻译的无参考后训练）[12:28] 🔍 InSight-doc: Agentic Visual Perception for Long-Document Understanding（InSight-doc：面向长文档理解的智能体视觉感知）[13:27] 🔀 UniMoMo: Expert Merging-Based MoE Acceleration for Large Recommendation Models（UniMoMo：基于专家合并的大型推荐模型MoE加速方法）【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

【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：[00:30] 🔄 Macaron-V1: Towards Open Continual Learning with Self-Improvement and Mixture-of-LoRA（Macaron-V1：迈向具备自我改进和LoRA混合的开放持续学习）[01:24] 🔧 SWE-Bench ProMax: Benchmarking Agents on Large-Scale Multilingual Code Refactoring（SWE-Bench ProMax：面向大规模多语言代码重构的智能体基准评测）[02:21] 🐍 Ouroboros: A Self-Developing Frontier Coding Agent with Reviewed Core Evolution（Ouroboros：通过核心评审进化实现自我发展的前沿编程智能体）[03:31] 🧠 BDH-CQ: In-Context Learning with Recurrent Latent Reasoning（BDH-CQ：基于循环潜在推理的上下文学习）[04:13] 🧠 Agent Memory Distillation: Empowering Small LLM Agents with Hierarchical Teacher Memory（智能体记忆蒸馏：利用分层教师记忆赋能小型大语言模型智能体）[05:11] 🧠 Motif 3: Technical Report（Motif 3：技术报告）[05:59] 🔬 Sci-VBench: Evaluating Knowledge- and Reasoning-Intensive Video Generation in Science Domains（Sci-VBench：评估科学领域中知识与推理密集型视频生成）[06:49] 🖼 What to Edit Next: Visually Aligned Image-Editing Follow-Up Suggestions in Conversational Systems（下一步编辑什么：对话系统中的视觉对齐图像编辑后续建议）[07:53] 🎯 SPOT: Sparse Probing and Outcome Calibration for On-Policy Distillation（SPOT：面向同策略蒸馏的稀疏探测与结果校准）[08:58] ⚡ OasisKV: Scaling In-Decode KV Cache Beyond HBM with Lookahead Sparse Prefetching（OasisKV：通过前瞻稀疏预取将解码期KV缓存扩展到HBM之外）[09:49] 🧠 RoMeRL: Balancing Feedback Coverage and the Memory-Reward Trap in Self-Evolving Agent Memory via Reduced-Order Utility States（RoMeRL：通过降阶效用状态平衡自进化智能体记忆中的反馈覆盖与记忆-奖励陷阱）[10:43] 🔍 Evidence-RL: Towards Evidence-intensive Visual Reasoning（证据强化学习：迈向证据密集型视觉推理）[11:42] 🧠 Scaling Inherently Interpretable Language Models（扩展内在可解释的语言模型）[12:40] 🧬 Evo-Bench: Can Language Models Improve Agent Harness?（Evo-Bench：语言模型能否改进智能体运行框架？）[13:40] 🔓 Stealing Reasoning Traces from Proprietary LLM APIs（从专有大语言模型API中窃取推理轨迹）【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

【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：[00:30] 🌍 Beyond Simply Environment Scaling: Designing Effective Environment Distributions for Multimodal Agent Learning（超越单纯的环境规模扩展：为多模态智能体学习设计有效的环境分布）[01:36] ⚔ SFT Conflicts, RL Coexists: A Theoretical and Empirical Analysis of Multi-Task Learning for LLMs（SFT冲突、RL共存：大语言模型多任务学习的理论与实证分析）[02:42] 🚗 SimWAM: A Simple World Action Model for End-to-End Autonomous Driving（SimWAM：面向端到端自动驾驶的简单世界动作模型）[03:40] 🎯 YOLO-PEFT: Parameter-Efficient Fine-Tuning on YOLO Family（YOLO-PEFT：YOLO系列上的参数高效微调）[04:43] 🎥 StreamArena: Toward Continuous, Interactive, and Long-Horizon Agentic Streaming Video Understanding（StreamArena：迈向连续、交互与长时程的智能体流式视频理解）[05:36] 🎧 Reinforcement Learning with Evolving Rubrics as Rewards for Audio Reasoning（以演化评分标准作为奖励的音频推理强化学习）[06:19] 🙈 When Activation Oracles Learn Not to Read: Concept-Specific Blind Spots in Fine-Tuned Oracles（当激活预言机学会不读取：微调预言机中的概念特异性盲区）[07:09] ⚡ Efficient Knowledge Distillation for LLMs: Offline Top-K Logits and a Fused Chunked KL Loss（面向大语言模型的高效知识蒸馏：离线Top-K Logits与融合分块KL损失）[08:16] 🚁 Uncertainty-Aware World Model for Aerial Image-Goal Navigation（面向航拍图像目标导航的不确定性感知世界模型）[09:19] 🎯 Douyin Multimodal Embedding Model Technical Report（抖音多模态嵌入模型技术报告）[10:19] 📈 Skaling: Chinchilla's Exponents Meet Kaplan's Coupling（Skaling：Chinchilla指数与Kaplan耦合的融合）[11:11] 🧠 Addressable Memory for Video World Models（视频世界模型的可寻址记忆）[12:05] 🧩 Relevant but Incomplete: Referential Dangling as a Paradigm-Level Failure Mode in Hard Prompt Compression（相关但不完整：硬提示压缩中作为范式级失败模式的指称悬空）[13:01] 🔄 Round-Trip Consistency: Bidirectional Diffusion Models Can Predict Their Own Rollout Errors（往返一致性：双向扩散模型可预测自身的展开误差）[14:01] 🧠 The Optimizer Is the Agent: Reasoning-Driven Search across Prompts, Programs, and ML Workflows（优化器即智能体：跨提示、程序与机器学习工作流的推理驱动搜索）【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

【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：[00:41] TOP1(🔥221) | 🔁 Recursive Synthesis for Long-Horizon Terminal Tasks（面向长时程终端任务的递归式合成）[03:39] TOP2(🔥162) | 🧭 LongHorizon-Harness: Advancing Long-Horizon Agents for Real-World Tasks（LongHorizon-Harness：推动面向真实世界任务的长时程智能体）[06:29] TOP3(🔥154) | 🎙 SwanTale: Unified Multi-Speaker Speech and Audio Generation for Instruct and Zero-Shot Tasks（SwanTale：面向指令与零样本任务的统一多说话人语音与音频生成）[09:28] TOP4(🔥140) | 🚗 Deferred Exposure of Future Trajectories for Verifiable Reasoning in Autonomous Driving VLMs（自动驾驶视觉语言模型中用于可验证推理的未来轨迹延迟暴露）[12:30] TOP5(🔥108) | ⚓ DAPD: Dual-Anchored Policy Distillation（双锚定策略蒸馏）【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

【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：[00:32] 🎯 AgentOPSD: Recursive Self-Distillation for Agentic Reinforcement Learning（AgentOPSD：面向智能体强化学习的递归自蒸馏）[01:35] 🤖 OSReward: Instituting Standardized Evaluation for Cross-Platform Computer-Use Reward Models（OSReward：为跨平台计算机使用奖励模型制定标准化评估）[02:24] 🌍 WorldClaw: Agentic 3D Open-World Generation at Scale（WorldClaw：大规模智能体式3D开放世界生成）[03:12] 🗺 GST-Bench: Can VLMs Develop Global Spatial Awareness from Video?（GST-Bench：视觉语言模型能否从视频中形成全局空间意识？）[04:17] 💭 EnvACE: Internalizing Environment Dynamics via World Rehearsal for Agentic Reinforcement Learning（EnvACE：通过世界预演将环境动态内化于智能体强化学习）[05:10] 🔍 Learning from Failures: Retrieval-Centric CoT via Hard Negatives for Unified Multimodal Retrieval（从失败中学习：基于硬负样本的检索中心思维链用于统一多模态检索）[06:08] ⏳ ChronoVision: Temporal Reasoning via Latent State Reconstruction（ChronoVision：通过潜在状态重建实现时序推理）[07:08] 🌐 From Economic Agents to Agentic Economies: A Systems Blueprint for Economic World Models（从经济主体到主体经济：经济世界模型的系统蓝图）[08:10] 🧮 On-Policy Delta Distillation for Multilingual Math Reasoning（面向多语言数学推理的同策略差值蒸馏）[08:59] ⚙ HarnessOpt-Bench: Evaluating LLMs at Harness Optimization（HarnessOpt-Bench：评估大语言模型在智能体运行框架优化上的表现）[09:43] 🇬 Teaching Nemotron Greek: Mining a Corpus, Adapting Retrieval, and Grounding Generation for Modern Greek across Specialist Domains（教Nemotron希腊语：面向专业领域的现代希腊语语料挖掘、检索适配与有据生成）[10:43] 🤖 DyPES-VLA: Learning Shared Dynamics Priors and Embodiment-Specific Control for Cross-Embodiment Manipulation（DyPES-VLA：学习共享动力学先验与具身特定控制以实现跨具身操作）[11:38] 🤖 World-to-Wrist: Task-Conditioned Future Wrist Modeling for Fine-Grained Robot Manipulation（世界到手腕：面向精细机器人操作的任务条件化未来手腕建模）[12:44] 📊 DataSpace: Benchmarking Data Agents for Verifiable Analytics over Heterogeneous Workspaces（DataSpace：面向异构工作空间的可验证分析数据代理基准测试）[13:44] 🪄 EffectLearner: World-Aware Object-Effect Reasoning for Real-World Video Object Removal（EffectLearner：面向真实世界视频目标移除的世界感知对象-效应推理）【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

【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：[00:32] 🔍 ABSeeker: Training Long-Horizon Search Agents via Answer-Backtracked Credit Assignment（ABSeeker：通过答案回溯的信用分配训练长程搜索智能体）[01:25] 🎨 ToolArtist: Tool-Using Unified Multimodal Models for Agentic Image Generation（ToolArtist：面向智能体图像生成的工具使用统一多模态模型）[02:16] 🪞 The Personalization Mirage: How LLMs Fabricate User Profiles, and Why Self-Monitoring Misleads（个性化幻象：大语言模型如何捏造用户画像，以及自我监控为何具有误导性）[03:17] ⚛ Towards Physics of Multimodal Pretraining: Knowledge Flow, Modality Synergy, Early Unification, and Recipes（迈向多模态预训练的物理机制：知识流动、模态协同、早期统一与配方）[04:20] 🤖 OneDayAgent: Towards a Long-Horizon Harness for Autonomous Agents（OneDayAgent：面向自主智能体的长周期任务执行框架）[05:14] 🧬 GDPevo: Evaluating Agent Self-Evolution on Real Business Tasks（GDPevo：在真实业务任务中评估智能体的自我进化）[06:05] 🎯 When Teachers Mislead: Spurious-Signal-Aware On-Policy Distillation（当教师误导：虚假信号感知的同策略蒸馏）[06:56] 🧩 Toward Skill-Native LLMs: Skill Entropy for Benchmarking and Training Long-Horizon Reasoning（迈向技能原生的大语言模型：用于长程推理评测与训练的技能熵）[08:02] 🧩 NOLLI: A Difficulty-Calibrated Puzzle Benchmark for Diagnosing the English-Korean Performance Gap（NOLLI：用于诊断英韩性能差距的难度校准谜题基准）[08:57] 🤖 Ego2Robot: Scalable Robot Data Synthesis from Egocentric Human Data（Ego2Robot：从第一人称人类数据中可扩展合成机器人数据）[09:54] 🎬 AVE-Compass: Towards Holistic Evaluation for Audio-Video Editing Abilities（AVE-Compass：面向音视频编辑能力的全面评估）[10:55] 🧠 When Memory Lies: An Empirical Study of Spatial Memory Staleness in VLM Agents（当记忆说谎：VLM智能体中空间记忆陈旧性的实证研究）[11:53] 🎯 Distill Where You Fail: Recovering Learning Signals of Negative RL-Groups from Adaptive Teacher Guidance（在失败处蒸馏：利用自适应教师指导恢复负强化学习组的学习信号）[12:49] 🧠 FocusMem: Factorizing Content, Readout, and Trust in Latent GUI Memory（FocusMem：分解潜在GUI记忆中的内容、读取与信任）[13:52] 👋 HelloWorld: Enabling Socially Interactive Characters in Video World Models（HelloWorld：在视频世界模型中实现社交互动角色）【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

【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] 🛒 MerchantBench: Benchmarking LLM Agents for Long-Term Coherence in E-Commerce Operations（MerchantBench：电商运营中LLM智能体长期一致性的基准评测）[01:19] 🎬 JoyAI-Video-Edit: Real-Time Open-Ended Video Editing with Autoregressive Diffusion（JoyAI-Video-Edit：利用自回归扩散的实时开放式视频编辑）[02:19] 🧊 Hunyuan3D-Buffalo 1.0: A Unified Multimodal Model for Scalable 3D Generation, Understanding, and Editing（Hunyuan3D-Buffalo 1.0：一种可扩展的三维生成、理解与编辑统一多模态模型）[03:15] 🌌 AURORA-LM: Autoencoding Unified Representation for Continuous-Latent Diffusion Language Modeling（AURORA-LM：自编码统一表示用于连续潜变量扩散语言建模）[04:18] 🎥 Video-DeepResearch: Towards the Next-Generation Multimodal Deepresearch Agent（视频深度研究：迈向下一代多模态深度研究智能体）[05:15] 🔄 Knowledge-Geometry Decoupling: Refreshable Pretrained Transfer for Streaming Recommendation（知识-几何解耦：面向流式推荐的可刷新预训练迁移）[06:20] 🤖 PCSD: Persistent Consistency for Self-Distillation in Agentic Reinforcement Learning（PCSD：智能体强化学习中自蒸馏的持久一致性）[07:08] 🌍 Quo Vadis, World Modeling?（世界建模，何去何从？）[07:59] 🔁 PAST-Bench: Benchmarking the Foundations of Recursive Self-Improvement in Personal Agents（PAST-Bench：对个人智能体中递归自我改进的基础进行基准测试）[08:56] 🌉 Any-OPD: Heterogeneous On-Policy Distillation for Flow-Matching Models via Representation-Space Bridging（Any-OPD：通过表示空间桥接实现异构流匹配模型的在策略蒸馏）[09:57] 🗜 OmniPack: Unified Token Compression for Efficient Omni-modal Large Language Models（OmniPack：用于高效全模态大语言模型的统一令牌压缩）[11:02] 📈 LLaDA MoE v2: Scaling Mixture-of-Experts Diffusion Language Models（LLaDA MoE v2：扩展混合专家扩散语言模型）[12:00] 🎯 CAPEval: A Decoupled Caption Evaluation across Understanding and Generation（CAPEval：面向理解与生成的解耦式图像描述评估）[12:55] 🧩 UniWorld-Design: From Pixel Generation to Layer-Native Design（UniWorld-Design：从像素生成到图层原生设计）[13:56] 🧠 SkillJack: Persistent Skill Backdoors in Self-Evolving Agents（SkillJack：自进化智能体中的持久性技能后门）【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

【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：[00:32] 🧭 LongHorizon-Harness: Advancing Long-Horizon Agents for Real-World Tasks（LongHorizon-Harness：推动面向真实世界任务的长时程智能体）[01:30] 🎙 SwanTale: Unified Multi-Speaker Speech and Audio Generation for Instruct and Zero-Shot Tasks（SwanTale：面向指令与零样本任务的统一多说话人语音与音频生成）[02:31] 🎯 VAD: Attributing Visual Evidence for Target Reconstruction in Multimodal On-Policy Distillation（VAD：在多模态在策略蒸馏中为目标重建归因视觉证据）[03:37] 🤖 Progressive Agent Skill Generation via Reinforcement Learning（基于强化学习的渐进式智能体技能生成）[04:39] ⚓ DAPD: Dual-Anchored Policy Distillation（双锚定策略蒸馏）[05:37] 🧲 UEmbed: Unified Sparse and Dense Multimodal Embeddings（UEmbed：统一稀疏与稠密多模态嵌入）[06:44] 🌍 WorldExam: Benchmarking World Models from Apparent Appearance to Inherent Reactivity（WorldExam：从表象外观到内在反应性的世界模型基准评测）[07:54] 🔗 CADENA: Stepwise CAD Reverse Engineering（CADENA：逐步式CAD逆向工程）[08:55] 🛠 SKT: Skill-Use Training at Scale via Verified Synthetic Data Generation（SKT：通过经验证的合成数据生成实现规模化技能使用训练）[09:58] 🤖 SWE-Touch: Benchmarking Coding Agents When Users Touch the Code（SWE-Touch：在用户改动代码时对编码代理的基准测试）[11:08] 🚗 Deferred Exposure of Future Trajectories for Verifiable Reasoning in Autonomous Driving VLMs（自动驾驶视觉语言模型中用于可验证推理的未来轨迹延迟暴露）[12:09] 🧠 WCM: A World Critic Model for Vision-Language-Action Reinforcement Learning（WCM：面向视觉-语言-动作强化学习的世界评论家模型）[13:05] 🔄 Motion Beyond Morphology: Bootstrapping Cross-Category Motion Transfer from Abstract Motion Representations（超越形态的运动：从抽象运动表征引导跨类别运动迁移）[14:04] 🧠 GradCuit: Credit-Assigned Gradient Flow Enables Robust and Interpretable Test-Time Latent Reasoning（GradCuit：信用分配的梯度流实现稳健且可解释的测试时潜在推理）[14:59] 🛋 Roomer: Reflective Object-Grounded Model Editing and Repair for 3D Indoor Layout Synthesis（Roomer：面向三维室内布局合成的反思式对象级模型编辑与修复）【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

【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：[00:42] TOP1(🔥473) | 🌍 Orca: The World is in Your Mind（虎鲸：世界在你心中）[03:50] TOP2(🔥438) | 🧠 Kimi K3: Open Frontier Intelligence（Kimi K3：开放前沿智能）[06:31] TOP3(🔥309) | 🧩 Program-as-Weights: A Programming Paradigm for Fuzzy Functions（程序即权重：面向模糊函数的编程范式）[09:09] TOP4(🔥307) | 🌍 ABot-World-0: Infinite Interactive World Rollout on a Single Desktop GPU（ABot-World-0：在单个桌面GPU上实现无限交互式世界展开）[12:34] TOP5(🔥293) | 🧪 AskChem: Claim-Centered Infrastructure for Chemistry Literature Synthesis（AskChem：以论断为中心的化学文献综合基础设施）[15:21] TOP6(🔥289) | 🤖 Qwen-UI-Agent Technical Report: Toward Next-Generation Real-World Centric Foundation GUI Agents（Qwen-UI-Agent技术报告：迈向下一代以真实世界为中心的基础GUI智能体）[18:00] TOP7(🔥259) | 🧠 Metis: Memory Foundation Model（Metis：记忆基础模型）[22:07] TOP8(🔥232) | 🧭 Harness Handbook: Making Evolving Agent Harnesses Readable,Navigable, and Editable（驾驭手册：使不断演化的智能体驾驭系统可读、可导航且可编辑）[25:09] TOP9(🔥205) | 🧠 LongStraw: Long-Context RL Beyond 2M Tokens under a Fixed GPU Budget（长稻草：在固定GPU预算下实现超过200万Token的长上下文强化学习）[28:24] TOP10(🔥198) | 🤖 RynnBrain 1.1: Towards More Capable and Generalizable Embodied Foundation Model（RynnBrain 1.1：迈向更强大和更通用的具身基础模型）【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

【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：[00:44] TOP1(🔥431) | 🧠 Kimi K3: Open Frontier Intelligence（Kimi K3：开放前沿智能）[03:17] TOP2(🔥292) | 🧪 AskChem: Claim-Centered Infrastructure for Chemistry Literature Synthesis（AskChem：以论断为中心的化学文献综合基础设施）[05:53] TOP3(🔥281) | 🤖 Qwen-UI-Agent Technical Report: Toward Next-Generation Real-World Centric Foundation GUI Agents（Qwen-UI-Agent技术报告：迈向下一代以真实世界为中心的基础GUI智能体）[08:46] TOP4(🔥256) | 🧠 Metis: Memory Foundation Model（Metis：记忆基础模型）[11:40] TOP5(🔥192) | 🤖 Progress Reward Modeling for Robotic Learning: A Comprehensive Survey（机器人学习的进度奖励建模：一项全面综述）【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

【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：[00:33] ⚡ TurboVLA: Real-Time Vision-Language-Action Model at 32 Hz on an RTX 4090 with <1 GB VRAM（TurboVLA：在RTX 4090上以32Hz频率运行且显存占用低于1GB的实时视觉-语言-动作模型）[01:36] 🎯 CoRT: Counterfactual Replay for Token-Level Rubric-Guided Policy Optimization（CoRT：用于令牌级准则引导策略优化的反事实重放）[02:31] 🤖 HumanCLAW: Can Vision-Language Models Act Through a Body?（HumanCLAW：视觉-语言模型能否通过身体行动？）[03:41] 🧬 DecoEvo: Score-Decoupled Co-Evolution of Solver and Rubric-Generator Skills in Text Space（DecoEvo：文本空间中求解器与评估生成器技能的解耦协同进化）[04:39] 🧠 CLBench-V: Evaluating Multimodal Context Learning from Grounding to Knowledge Acquisition（CLBench-V：从基础定位到知识获取的多模态上下文学习评估）[05:32] 🧩 CAST: Game Solvers as Turn-Level Teachers for LLM Agents（CAST：游戏求解器作为LLM智能体的回合级教师）[06:28] 🧠 SkillRise: Agentic Reinforcement Learning for Cross-Task Skill Evolution（技能崛起：面向跨任务技能演化的智能体强化学习）[07:18] 🎮 StatePlay: State-Aware Game World Models for Mechanics-Consistent Generation（StatePlay：状态感知的游戏世界模型用于机制一致的内容生成）[08:05] 📊 OmegaUse-OfficeVal: Benchmarking LLM Agents on Long-Horizon Office-Suite Tasks with Economic Grounding（OmegaUse-OfficeVal：基于经济基准评估LLM智能体在长期办公套件任务中的表现）[09:08] 🤖 Can AI agents conduct open-ended AI research? Early evidence from two case studies（AI智能体能否进行开放式的AI研究？来自两个案例研究的早期证据）[10:04] 🛡 GPT-Red: Automated Red Teaming via Self-Play at Scale（GPT-Red：通过大规模自我对弈实现自动化红队测试）[11:04] 🎬 Explicit Layer Modeling for Video Object Insertion and Layer Decomposition（显式层建模用于视频对象插入与层分解）[12:01] 🕵 StealthBench: Measuring Operational Stealth in Autonomous Offensive-Security Agents（StealthBench：衡量自主攻击安全代理的操作隐蔽性）[12:54] 🧠 Memory for Large Language Models（大型语言模型的记忆机制）[13:48] 📜 Grading the Narrators: An Isnad-Rijal Framework for Claim-Level Provenance in Multi-Agent Knowledge Systems（为叙述者评级：面向多智能体知识系统中声明级溯源的一种伊斯纳德-里贾尔框架）【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

【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：[00:31] 🤖 HiFi-UMI: Learning Deployable Manipulation Policies from High-Fidelity UMI Data Alone（HiFi-UMI：仅从高保真UMI数据学习可部署的操作策略）[01:31] 🔍 A New Role for Relevance: Guiding Corpus Interaction in Agentic Search（相关性的新角色：在智能体搜索中引导语料库交互）[02:20] 🎨 ReDesign: Recovering Editable Design Structures from Images via Agentic Decomposition（ReDesign：通过智能体分解从图像中恢复可编辑的设计结构）[03:07] 🧠 Keep It InMind: Benchmarking the Implicit-Association Blind Spot in Agent Memory（保持铭记：基准测试代理记忆中的内隐关联盲点）[03:56] 🏃 Pass the Baton: Trajectory-Relayed On-Policy Distillation（传递接力棒：轨迹中继的在线策略蒸馏）[04:47] ⚡ Mage-VL: An Efficient Codec-Native Streaming Multimodal Foundation Model（Mage-VL：一种高效的编解码器原生流式多模态基础模型）[05:49] 🧩 CodeNib: A Multi-View Data System for Serving Repository Context to Coding Agents（CodeNib：一种为编码智能体提供仓库上下文服务的多视图数据系统）[06:48] 🌍 Wonder: Video World Model Done Better（Wonder：更优的视频世界模型）[07:51] 👁 PerceptionBench: Evaluating Atomic Visual Perception in Multimodal Large Language Models（感知基准：评估多模态大语言模型中的原子视觉感知能力）[08:44] 🔍 Novel Claim or Déjà Vu? Rethinking "Contamination-Free'' Dynamic Evaluation for Multimodal Automated Fact-Checking（新颖主张还是似曾相识？重新思考多模态自动事实核查的“无污染”动态评估）[09:40] 🛡 Shieldstral（盾星）[10:41] 🔀 MODUS: Decoder-Only Any-to-Any Modeling of Diverse Modalities（MODUS：仅解码器的任意模态到任意模态多样化建模）[11:41] ⚡ Parallel Decoding Distillation for Fast Image and Video Generation（并行解码蒸馏：面向快速图像与视频生成的方法）[12:23] 🎬 Visual prompt engineering for video models（视频模型的视觉提示工程）[13:16] 🎯 OmniDelta: Skill-Driven Budget Allocation for Token Compression in OmniLLMs（OmniDelta：面向全模态大语言模型令牌压缩的技能驱动预算分配方法）【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

【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：[00:33] 🧠 Kimi K3: Open Frontier Intelligence（Kimi K3：开放前沿智能）[01:22] 🎨 JarvisHub: An Open Harness for Canvas-Native Multimodal Creative Agents（JarvisHub：一个面向画布原生多模态创意代理的开放框架）[02:20] 🤖 From Proprietary to Open-Source: Bridging the Distribution Gap via Multi-Agent Protocol Distillation in Agentic Search（从专有到开源：通过多智能体协议蒸馏弥合智能搜索中的分布差距）[03:18] 🤖 Progress Reward Modeling for Robotic Learning: A Comprehensive Survey（机器人学习的进度奖励建模：一项全面综述）[04:17] 🤖 StateAct: Program State, before Pixels, for Long-Horizon Computer-Use Agents（StateAct：面向长周期计算机使用代理，程序状态优先于像素）[05:23] 🧠 Rethinking Classifier-Free Guidance in On-Policy Diffusion Distillation（重新思考在线策略扩散蒸馏中的无分类器引导）[06:27] 🗼 Data Pyramid for Embodied Manipulation（具身操作的数据金字塔）[07:33] ⚡ Sol-Attn: Accelerating Video Generation Inference via On-the-Fly Attention Sparsification（Sol-Attn：通过即时注意力稀疏化加速视频生成推理）[08:37] 🎬 OmniVAE: An Audio-Video VAE with Cross-Modal Alignment for Joint Generation（OmniVAE：一种具有跨模态对齐的音频-视频VAE，用于联合生成）[09:30] 🧠 The Physics of Multi-Turn Long-Horizon Planning: From Pre-training to Post-training via Single- and Multi-Teacher On-Policy Agentic Distillation（多轮长程规划中的物理学：通过单教师与多教师在线策略智能体蒸馏从预训练到后训练）[10:26] 👗 Oxygen-TryOn: Fashion-Native Foundation Model for Any-item Virtual Try-On（氧气试穿：面向时尚的原生基础模型，实现任意物品虚拟试穿）[11:17] 🦎 Chamaileon: Cross-Context Binder Design with Contextualized Modeling and Mixed Sampling（Chamaileon：基于情境化建模与混合采样的跨情境结合剂设计）[12:11] 🔮 dRAE: Representation Autoencoder with Hyper-Spherical Codes（dRAE：超球面码的表示自编码器）[13:09] 🧩 DecoupleMix: Decoupled Ratio Search and Convex Allocation for Scalable VLM Data Recipes（解耦混合：用于可扩展VLM数据配方的解耦比率搜索与凸分配）[14:08] 🏥 ClinFusion: A Vision-Centric Multimodal LLM System for Holistic Medical Understanding（ClinFusion：一种面向整体医学理解的以视觉为中心的多模态大语言模型系统）【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

【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 14 篇论文如下：[00:32] 🔄 Skill Self-Play: Pushing the Frontier of LLM Capability with Co-Evolving Skills（技能自我对弈：通过协同进化技能推动大语言模型能力前沿）[01:22] 🧠 Agentic Context Management: Solving Agent Memory and Cost by Treating Them as Lifecycle and Architecture Problems（智能体上下文管理：将代理记忆与成本视为生命周期与架构问题）[02:18] 🤖 Molt: A Scalable PyTorch-Native Training Framework for Agentic Reinforcement Learning（Molt：一个可扩展的、原生PyTorch的智能体强化学习训练框架）[03:07] 🧪 DataPrep-Bench: Benchmarking LLMs as Training Data Preparators（数据准备基准：将大语言模型作为训练数据准备工具的基准测试）[04:04] 📊 Scaling Native Multimodal Pre-Training From Scratch（从头开始扩展原生多模态预训练）[04:54] 🎯 Three-Body Scattering for Generative Modeling（用于生成建模的三体散射）[05:48] 🌐 LAMAR: An Open Language-Aware Multilingual Alignment Reranker（LAMAR：一种开放的语言感知多语言对齐重排序器）[06:46] 🧠 Multi-Head Latent Control: A Unified Interface for LLM Agent Decision Making（多头潜在控制：大语言模型智能体决策的统一接口）[07:43] 🎛 Spectral Prior for Reducing Exposure Bias in Diffusion Models（用于减少扩散模型曝光偏差的频谱先验）[08:34] 🧠 IDEAgent: Agentic Quality-Diversity Search for Research Idea Generation（IDEAgent：面向研究创意生成的主体性质量-多样性搜索）[09:27] 🎯 SceneActBench: Can Agents Act on the 3D Scenes They See?（场景动作基准：智能体能否对所见的三维场景采取行动？）[10:37] 🔄 Closing the Loop: Training-Free Revisit Consistency for Autoregressive Generative Rendering（闭环：无需训练的循环一致性自回归生成渲染）[11:31] 🔊 Multimodal Speaker Verification as a Threat to Speaker Anonymization（多模态说话人验证对说话人匿名化的威胁）[12:26] 🧠 VisCo: Leveraging Large Language Models as Intrinsic Encoders for Visual Token Compression（VisCo：利用大语言模型作为视觉标记压缩的内在编码器）【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

【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：[00:47] TOP1(🔥297) | 🌍 ABot-World-0: Infinite Interactive World Rollout on a Single Desktop GPU（ABot-World-0：在单个桌面GPU上实现无限交互式世界展开）[02:39] TOP2(🔥197) | 🤖 RynnBrain 1.1: Towards More Capable and Generalizable Embodied Foundation Model（RynnBrain 1.1：迈向更强大和更通用的具身基础模型）[04:48] TOP3(🔥165) | ⏱ TimeLens2: Generalist Video Temporal Grounding with Multimodal LLMs（TimeLens2：基于多模态大语言模型的通用视频时间定位）[06:35] TOP4(🔥144) | 🔍 RAGU: A Multi-Step GraphRAG Engine with a Compact Domain-Adapted LLM（RAGU：一种具有紧凑领域自适应大语言模型的多步图检索增强生成引擎）[09:10] TOP5(🔥141) | 🔍 AREX: Towards a Recursively Self-Improving Agent for Deep Research（AREX：迈向递归自我改进的深度研究智能体）【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

【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：[00:32] 🔍 AREX: Towards a Recursively Self-Improving Agent for Deep Research（AREX：迈向递归自我改进的深度研究智能体）[01:34] 🤖 ReferTrack: Referring Then Tracking for Embodied Visual Tracking（ReferTrack：面向具身视觉追踪的“先指认后跟踪”范式）[02:20] 📚 K12-KGraph: A Curriculum-Aligned Knowledge Graph for Benchmarking and Training Educational LLMs（K12-KGraph：一个面向课程对齐的知识图谱，用于基准测试和训练教育大语言模型）[03:13] 🖼 Visual Contrastive Self-Distillation（视觉对比自蒸馏）[04:02] 🗺 Show, Don't Tell: Evaluating Spatial Cognition in Generative Pixels Rather Than LLM Text（展示而非叙述：在生成像素而非LLM文本中评估空间认知）[04:51] 🎨 Color Pass-Through via Camera-Display Coupling（通过相机-显示耦合的色彩直通）[05:45] 🛠 Tencent WorkBuddy Bench: A Multi-Domain Coding-Agent Benchmark with Contamination-Resistant Task Construction（腾讯工作伙伴基准：一个具有抗污染任务构建的多领域编码智能体基准）[06:43] 🧭 LLMs Get Lost in Evolving User Intent（大语言模型在用户意图演变中迷失方向）[07:33] 🎥 Self-Supervised Learning of Structured Dynamics from Videos（从视频中自监督学习结构化动力学）[08:28] 🧠 Sample-Efficient Learning from Agent Experience（从智能体经验中进行样本高效学习）[09:28] 🌀 Recurrent Sinusoidal INRs for Efficient High-Fidelity Representation（用于高效高保真表示的递归正弦隐式神经表示）[10:27] 🌍 Streaming Multi-Agent Autoregressive Diffusion Model with World State Registers（流式多智能体自回归扩散模型与世界状态寄存器）[11:28] 🤖 Robostral Navigate（罗博斯特拉导航）[12:18] 🎭 Predictive Divergence Masks for LLM RL（预测性散度掩码用于大语言模型强化学习）[13:04] 🎬 SANA-Video 2.0: Hybrid Linear Attention with Attention Residuals for Efficient Video Generation（SANA-Video 2.0：混合线性注意力与注意力残差实现高效视频生成）【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

【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：[00:34] 🎮 Generative World Renderer at the Speed of Play（以游戏速度运行的生成式世界渲染器）[01:27] 🔀 DataFlow-Harness: A Grounded Code-Agent Platform for Constructing Editable LLM Data Pipelines（数据流驾驭平台：一种用于构建可编辑LLM数据管道的接地代码代理平台）[02:16] 🔍 Text Template Tokens Are Implicit Semantic Registers in Diffusion Transformers（文本模板标记是扩散Transformer中的隐式语义寄存器）[03:04] ⚡ Mage-Flow: An Efficient Native-Resolution Foundation Model for Image Generation and Editing（Mage-Flow：一种用于图像生成与编辑的高效原生分辨率基础模型）[03:52] 🌍 AlayaWorld: Interactive Long-Horizon World Modeling -- Full Technical Report（AlayaWorld：交互式长时域世界建模——完整技术报告）[04:49] 🤖 Stale but Stable: Staleness-Adaptive Trust Regions for Stabilizing Asynchronous Reinforcement Learning（陈旧但稳定：面向异步强化学习稳定化的陈旧性自适应信任区域方法）[05:52] 🌍 ABot-World-0: Infinite Interactive World Rollout on a Single Desktop GPU（ABot-World-0：在单个桌面GPU上实现无限交互式世界展开）[06:49] 📊 SciForma: Structure-Faithful Generation of Scientific Diagrams（SciForma：科学图表的结构忠实生成）[07:45] 🔍 AgentDebugX: An Open-Source Toolkit for Failure Observability, Attribution, and Recovery in LLM Agents（AgentDebugX：用于LLM Agent故障可观测性、归因与恢复的开源工具包）[08:41] ⚡ HPD-Parsing: Hierarchical Parallel Document Parsing（HPD-Parsing：层级并行文档解析）[09:38] 📊 Two-Level Meta-Rubrics for Evaluating Open-Ended Generation: GAMUT, a Benchmark for Factual Completeness（用于评估开放生成的两级元评价标准：GAMUT，一个面向事实完整性的基准测试）[10:33] 🧠 ISO: An RLVR-Native Optimization Stack（ISO：一种原生于RLVR的优化栈）[11:41] 🎙 Transcription Policy as a Latent Variable: Activating Controllable Verbatim ASR with Word-Level Timing（转录策略作为潜在变量：通过词级时序激活可控的逐字语音识别）[12:43] 🎓 EduPanel: A Three-Agent LLM Judge for Teaching Videos -- Reliability, Complementarity, and Human Trust Calibration（EduPanel：一种用于教学视频的三智能体LLM评审器——可靠性、互补性与人类信任校准）[13:29] 🎬 Masked Visual Actions for Unified World Modeling（掩蔽视觉动作：用于统一世界建模）【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

【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：[00:31] ⏱ TimeLens2: Generalist Video Temporal Grounding with Multimodal LLMs（TimeLens2：基于多模态大语言模型的通用视频时间定位）[01:14] ✂ SWE-Pruner Pro: The Coder LLM Already Knows What to Prune（SWE-Pruner Pro：编码器大语言模型已知道该剪枝什么）[02:05] 🌍 EvolvingWorld: An Open-Schema Framework for Co-Evolving Role-Play Agents and World Model in Interactive Literary World（演化世界：面向交互式文学世界中角色与世界观协同演化的开放模式框架）[03:08] 🔍 DeepSearch-World: Self-Distillation for Deep Search Agents in a Verifiable Environment（DeepSearch-World：可验证环境中深度搜索代理的自我蒸馏）[04:06] 🎬 HOMIE: Human-object Centric Video Personalization via Multimodal Intelligent Enchancement（HOMIE：通过多模态智能增强实现以人-物为中心的视频个性化）[04:59] 🤖 RynnBrain 1.1: Towards More Capable and Generalizable Embodied Foundation Model（RynnBrain 1.1：迈向更强大和更通用的具身基础模型）[05:51] 🍎 Apple-$π$: Benchmarking Thinking with Video Towards Law-Grounded Physical Intelligence（苹果-π：面向法律约束的物理智能，以视频为基准测试思维过程）[06:49] 🎯 Group Entropy-Controlled Policy Optimization（群体熵控制策略优化）[07:52] 🧠 ReflectWorld-MM: An Entity-Oriented Multimodal Memory System for Open-Ended Video Streams（反射世界-多模态：面向开放视频流的实体导向多模态记忆系统）[09:03] 🌍 GigaAM Multilingual: Foundation Model for Underrepresented Languages（GigaAM多语言模型：面向低资源语言的基座模型）[09:51] ⏱ GigaChat Audio: Time-aware Large Audio Language Model（GigaChat音频：时间感知的大规模音频语言模型）[10:50] 🤖 Environment-free Synthetic Data Generation for API-Calling Agents（面向API调用智能体的无环境合成数据生成）[11:42] 🎬 FlowMimic: Mask-free Visual Editing and Generation with Pixel-pair Warped Flow Field for Online Video Editing Data Generation and Modality Mimicry（FlowMimic: 基于像素对扭曲流场的无掩码视觉编辑与生成——用于在线视频编辑数据生成与模态模仿）[12:30] 🧊 DiffGI: Differentiable Geometry Images for High-Fidelity Thin-Shell 3D Generation（DiffGI：面向高保真薄壳三维生成的可微几何图像）[13:20] 🎓 LLM-as-a-Coach: Experiential Learning for Non-Verifiable Tasks（LLM作为教练：非可验证任务的体验式学习）【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

【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：[00:32] 🔍 RAGU: A Multi-Step GraphRAG Engine with a Compact Domain-Adapted LLM（RAGU：一种具有紧凑领域自适应大语言模型的多步图检索增强生成引擎）[01:35] 🎯 RESOURCE2SKILL: Distilling Executable Agent Skills from Human-Created Multimodal Resources（RESOURCE2SKILL：从人类创建的多模态资源中提炼可执行的智能体技能）[02:34] 🤖 Xiaomi-Robotics-1: Scaling Vision-Language-Action Models with over 100K Hours of Real-World Trajectories（小米机器人-1：基于超过10万小时真实世界轨迹数据扩展视觉-语言-动作模型）[03:21] 🔗 xHC: Expanded Hyper-Connections（xHC：扩展超连接）[04:09] 🔄 Loop the Loopies!（循环循环！）[05:01] 🏥 Cura 1T: Specialized Model for Agentic Healthcare（Cura 1T：面向智能体医疗的专用模型）[05:47] 🧠 RecGPT-V3 Technical Report（RecGPT-V3技术报告）[06:34] 🧪 On-Policy Delta Distillation（在策略增量蒸馏）[07:20] 🤖 From Human-Centric to Agentic Code Review: The Impact of Different Generations of Generative AI Technology on Review Quality（从以人为中心到智能体代码审查：不同代际生成式人工智能技术对审查质量的影响）[08:14] 🎵 Qwen-Music Technical Report（Qwen-Music技术报告）[09:08] ♟ Understanding Reasoning from Pretraining to Post-Training（理解从预训练到后训练中的推理能力）[10:00] 🤖 When Does Muon Help Agentic Reinforcement Learning?（缪子优化器何时有助于智能体强化学习？）[10:56] 🔄 Recursive Harness Self-Improvement（递归式框架自我改进）[11:50] 🎥 VideoRAE: Taming Video Foundation Models for Generative Modeling via Representation Autoencoders（VideoRAE：通过表示自编码器驯服视频基础模型以进行生成建模）[12:49] 🦩 Audio-Visual Flamingo: Open Audio-Visual Intelligence for Long and Complex Videos（音频-视觉火烈鸟：面向长视频与复杂视频的开放音频-视觉智能）【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

【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：[00:52] TOP1(🔥200) | 🧭 Harness Handbook: Making Evolving Agent Harnesses Readable,Navigable, and Editable（驾驭手册：使不断演化的智能体驾驭系统可读、可导航且可编辑）[03:22] TOP2(🔥179) | 🧠 LongStraw: Long-Context RL Beyond 2M Tokens under a Fixed GPU Budget（长稻草：在固定GPU预算下实现超过200万Token的长上下文强化学习）[06:16] TOP3(🔥150) | 🎥 VideoChat3: Fully Open Video MLLM for Efficient and Generalist Video Understanding（VideoChat3：面向高效与通用视频理解的全开放视频多模态大语言模型）[08:55] TOP4(🔥130) | 🧠 Weak-to-Strong Generalization via Direct On-Policy Distillation（弱到强泛化：通过直接在线策略蒸馏）[11:09] TOP5(🔥127) | 🖼 Boogu-Image-0.1: Boosting Open-Source Unified Multimodal Understanding and Generation（Boogu-Image-0.1：推动开源统一多模态理解与生成）【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

【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：[00:33] 🧭 Harness Handbook: Making Evolving Agent Harnesses Readable,Navigable, and Editable（驾驭手册：使不断演化的智能体驾驭系统可读、可导航且可编辑）[01:19] 🖼 Boogu-Image-0.1: Boosting Open-Source Unified Multimodal Understanding and Generation（Boogu-Image-0.1：推动开源统一多模态理解与生成）[02:07] 🧠 Ring-Zero: Scaling Zero RL to a Trillion Parameters for Emergent Reasoning（环零：将零强化学习扩展至万亿参数以实现涌现推理）[03:08] 📄 OvisOCR2 Technical Report（OvisOCR2 技术报告）[04:12] 🤖 KnowAct-GUIClaw: Know Deeply, Act Perfectly, Personal GUI Assistant with Self-Evolving Memory and Skill（知深行准-GUIClaw：具备自进化记忆与技能的深度认知、精准执行个人GUI助手）[05:10] 🛡 PolicyShiftGuard: Benchmarking and Improving Policy-Adaptive Image Guardrails（政策转变守卫：基准测试与改进策略自适应图像防护机制）[06:02] 🤖 GigaWorld-Policy-0.5: A Faster and Stronger WAM Empowered by AutoResearch（GigaWorld-Policy-0.5：一种由自动研究驱动的更快更强的世界动作模型）[06:57] 🖼 MetaView: Monocular Novel View Synthesis with Scale-Aware Implicit Geometry Priors（MetaView：具有尺度感知隐式几何先验的单目新视角合成）[07:46] ✂ ShortOPD: Recovering Pruned LLMs with Short-to-Long On-Policy Distillation（ShortOPD：通过短到长在线策略蒸馏恢复剪枝后的大语言模型）[08:40] 🎭 Hallo4D: Multi-Modal Hallucination Mitigation for Consistent Spatio-Temporal Generation（Hallo4D：多模态幻觉缓解实现一致的时空生成）[09:28] 🔬 Registers Matter for Pixel-Space Diffusion Transformers（寄存器对像素空间扩散Transformer至关重要）[10:22] 🤖 Vinci2: Providing Proactive Assistance in Continuous Egocentric Videos（Vinci2：在连续自我中心视频中提供主动辅助）[11:21] 🔍 Tracing Agentic Failure from the Flow of Success（从成功流程中追溯智能体失败）[12:20] 🔍 From Noisy Traces to Root Causes: Structural Trajectory Analysis and Causal Extraction for Agent Optimization（从噪声轨迹到根因：面向智能体优化的结构化轨迹分析与因果提取）[13:23] 🧭 AgentCompass: A Unified Evaluation Infrastructure for Agent Capabilities（AgentCompass：一种统一的智能体能力评估基础设施）【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

【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：[00:31] 🔄 Read It Back: Pretrained MLLMs Are Zero-Shot Reward Models for Text-to-Image Generation（重新读它：预训练多模态大语言模型是文本到图像生成的零样本奖励模型）[01:35] 🔍 Blind-Spots-Bench: Evaluating Blind Spots in Multimodal Models（盲点基准：评估多模态模型中的盲点）[02:30] 📄 SynthDocBench: Controlled Benchmark for Long-Context Visual Document Understanding（SynthDocBench：面向长上下文视觉文档理解的受控基准）[03:31] 🔍 Know Before Fix: QA-Driven Repository Knowledge Acquisition for Software Issue Resolution（先知晓再修复：面向软件问题修复的基于问答的仓库知识获取）[04:20] 🎵 MuScriptor: An Open Model for Multi-Instrument Music Transcription（MuScriptor：面向多乐器音乐转录的开放模型）[05:12] 🔍 Search Beyond What Can Be Taught: Evolving the Knowledge Boundary in Agentic Visual Generation（超越可教授的知识边界：在智能体视觉生成中演化知识边界）[06:00] 🎨 Let RGB Be the Language of Vision（让RGB成为视觉的语言）[06:50] 📄 MonkeyOCRv2: A Visual-Text Foundation Model for Document AI（MonkeyOCRv2：面向文档AI的视觉-文本基础模型）[07:50] 🤖 Towards Autonomous and Auditable Medical Imaging Model Development（迈向自主且可审计的医学影像模型开发）[08:40] 🧠 Principled Analysis of Deep Reinforcement Learning Evaluation and Design Paradigms（深度强化学习评估与设计范式的原则性分析）【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

【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 14 篇论文如下：[00:32] 🧠 Weak-to-Strong Generalization via Direct On-Policy Distillation（弱到强泛化：通过直接在线策略蒸馏）[01:22] 🤖 ABot-N1: Toward a General Visual Language Navigation Foundation Model（ABot-N1：迈向通用视觉语言导航基础模型）[02:20] 🤖 ABot-AgentOS: A General Robotic Agent OS with Lifelong Multi-modal Memory（ABot-AgentOS：一种具有终身多模态记忆的通用机器人智能体操作系统）[03:21] 🕺 4D Human-Scene Reconstruction from Low-Overlap Captures（低重叠度捕获下的四维人体场景重建）[04:13] 🧠 LightMem-Ego: Your AI Memory for Everyday Life（轻量记忆自我：你的日常生活AI记忆）[05:15] 🧮 AdvancedMathBench: A Benchmark Suite for Advanced Mathematical Proof Generation and Verification（高级数学推理基准：面向高级数学证明生成与验证的基准套件）[06:11] 🧠 Metacognition in LLMs: Foundations, Progress, and Opportunities（大语言模型中的元认知：基础、进展与机遇）[07:03] 🤖 EgoSteer: A Full-Stack System Towards Steerable Dexterous Manipulation from Egocentric Videos（EgoSteer：面向从自我中心视频实现可操控灵巧操作的全栈系统）[07:55] 🧩 Proxy Exploration and Reusable Guidance: A Modular LLM Post-Training Paradigm via Proxy-Guided Update Signals（代理探索与可复用引导：一种通过代理引导更新信号实现的模块化大语言模型后训练范式）[08:55] 🧠 NeuroCogMap Reveals Cognitive Organization of Large Language Models（神经认知图谱揭示大型语言模型的认知组织）[09:52] 🎬 Motion4Motion: Motion Transfer Across Subjects at Inference（Motion4Motion：推理时跨主体的运动迁移）[10:58] 👗 CtrlVTON: Controllable Virtual Try-On via Visual-Instance-Prompt Segmentation（CtrlVTON：通过视觉实例提示分割实现可控虚拟试穿）[11:54] 🎭 Latent-Identity Tuning in Text-to-Image Personalization Models（文本到图像个性化模型中的潜在身份调优）[12:52] 🧊 LATO.2: Factorized 3D Mesh Generation with Vertex and Topology Flow（LATO.2：基于顶点与拓扑流的分解式3D网格生成）【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

【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 13 篇论文如下：[00:33] 🚀 Long-Horizon-Terminal-Bench: Testing the Limits of Agents on Long-Horizon Terminal Tasks with Dense Reward-Based Grading（长时终端基准：在具有密集奖励评分的长期终端任务中测试智能体的极限）[01:25] 👁 Scalable Visual Pretraining for Language Intelligence（面向语言智能的可扩展视觉预训练）[02:22] 🎥 Video Generation Models are General-Purpose Vision Learners（视频生成模型是通用视觉学习器）[03:16] 🎯 Trust Region Policy Distillation（信任区域策略蒸馏）[04:14] 🧠 KronQ: LLM Quantization via Kronecker-Factored Hessian（KronQ：通过Kronecker分解黑塞矩阵实现的大语言模型量化）[05:00] 🌍 PanoWorld: Real-World Panoramic Generation（PanoWorld：真实世界全景生成）[05:56] 🎨 From RGB Generation to Dense Field Readout: Pixel-Space Dense Prediction with Text-to-Image Models（从RGB生成到密集场读取：利用文本到图像模型进行像素空间密集预测）[06:49] 🎯 Self-Guided Test-Time Training for Long-Context LLMs（自引导测试时训练用于长上下文大语言模型）[07:36] 🚗 Flow-ERD: Agent-type Aware Flow Matching with Entropy-Regularized Distillation for Diverse Traffic Simulation（Flow-ERD：面向多样化交通仿真的智能体类型感知流匹配与熵正则化蒸馏）[08:28] 🔊 Phone Segmentation and Recognition through Phonological Activation Mapping（通过音位激活映射进行音素分割与识别）[09:28] 🧠 Towards Mechanistically Understanding Why Memorized Knowledge Fails to Generalize in Large Language Model Finetuning（探究大语言模型微调中记忆知识为何无法泛化的机制性理解）[10:18] 🤖 A Sovereign, Open-Source Foundation Model for German and English（一个面向德语和英语的主权开源基础模型）[11:13] 🏛 VaseMuseum: Digital Intelligent Museum for Ancient Greek Pottery（VaseMuseum：面向古希腊陶器的数字智能博物馆）【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

【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：[00:51] TOP1(🔥161) | 🎯 The Mirage of Optimizing Training Policies: Monotonic Inference Policies as the Real Objective for LLM Reinforcement Learning（优化训练策略的幻影：单调推理策略作为大语言模型强化学习的真正目标）[02:56] TOP2(🔥119) | 🎬 Vidu S1: A Real-Time Interactive Video Generation Model（Vidu S1：一种实时交互式视频生成模型）[04:56] TOP3(🔥89) | 🤖 RynnWorld-4D: 4D Embodied World Models for Robotic Manipulation（RynnWorld-4D：面向机器人操作的4D具身世界模型）[07:15] TOP4(🔥84) | 🎮 AlayaWorld: Long-Horizon and Playable Video World Generation（AlayaWorld：长视界且可游玩的视频世界生成）[09:58] TOP5(🔥83) | 🔬 Accurate, Interdisciplinary and Transparent Structure-property Understanding with Deep Native Structural Reasoning（基于深度原生结构推理的精确、跨学科且透明的构效关系理解）【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

【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：[00:31] 🎬 Vidu S1: A Real-Time Interactive Video Generation Model（Vidu S1：一种实时交互式视频生成模型）[01:29] 🔍 Video-Oasis: Rethinking Evaluation of Video Understanding（视频绿洲：重新思考视频理解的评估）[02:27] 🔑 Why Can't I Open My Drawer? Mitigating Object-Driven Shortcuts in Zero-Shot Compositional Action Recognition（为什么我打不开抽屉？缓解零样本组合动作识别中的对象驱动捷径）[03:21] 🤖 UniClawBench: A Universal Benchmark for Proactive Agents on Real-World Tasks（UniClawBench：面向真实世界任务中主动智能体的通用基准）[04:18] 🎥 LongE2V: Long-Horizon Event-based Video Reconstruction, Prediction, and Frame Interpolation with Video Diffusion Models（LongE2V：基于视频扩散模型的长时域事件驱动视频重建、预测与帧插值）[05:24] 🧬 Ideas Have Genomes: Benchmarking Scientific Lineage Reasoning and Lineage-Grounded Idea Generation（思想具有基因组：科学谱系推理与基于谱系的创意生成基准测试）[06:13] 🌐 Enhancing In-context Panoramic Generation via Geometric-aware Pretraining（通过几何感知预训练增强上下文全景生成）[06:54] 🎬 CineMobile: On-Device Image-to-Video Diffusion for Cinematic Camera Motion Generation（CineMobile：面向电影级摄像机运动生成的设备端图像到视频扩散）[07:52] 🎬 OpenCoF: Learning to Reason Through Video Generation（OpenCoF：通过视频生成学习推理）[08:45] 🚀 Jet-Long: Efficient Long-Context Extension with Dynamic Bifocal RoPE（Jet-Long：基于动态双焦点旋转位置编码的高效长上下文扩展）[09:39] ⚡ Linear Attention Architectures: Mechanisms, Trade-offs, and Cross-Layer Routing（线性注意力架构：机制、权衡与跨层路由）[10:38] 💊 DrugGen 2: A disease-aware language model for enhancing drug discovery（DrugGen 2：一种疾病感知型语言模型，用于增强药物发现）[11:23] ⚡ Flash-BoN: Instant Drafts for Inference-Time Scaling in Diffusion Models（Flash-BoN：扩散模型中推理时缩放的即时草稿生成）[12:08] 🎵 A Quantized Native Runtime for On-Device Semantic Audio Generation（面向设备端语义音频生成的量化原生运行时）[13:01] 🚀 UP: Unbounded Positive Asymmetric Optimization for Breaking the Exploration-Stability Dilemma（UP：用于打破探索-稳定性困境的无界正向非对称优化）【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

【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 7 篇论文如下：[00:30] 🔬 Accurate, Interdisciplinary and Transparent Structure-property Understanding with Deep Native Structural Reasoning（基于深度原生结构推理的精确、跨学科且透明的构效关系理解）[01:28] 🤖 Dual Latent Memory in Vision-Language-Action Models for Robotic Manipulation（视觉-语言-动作模型中的双潜记忆用于机器人操作）[02:34] 🤖 Scaling Mixture-of-Experts Video Pretraining for Embodied Intelligence（面向具身智能的混合专家视频预训练规模化）[03:28] 🌍 Infinite Worlds with Versatile Interactions（无限世界与多样化交互）[04:32] 🤖 RoboDojo: A Unified Sim-and-Real Benchmark for Comprehensive Evaluation of Generalist Robot Manipulation Policies（RoboDojo：统一仿真与真实世界的通用机器人操作策略综合评估基准）[05:34] 🏙 WildCity: A Real-World City-Scale Testbed for Rendering, Simulation, and Spatial Intelligence（WildCity：一个用于渲染、仿真和空间智能的真实世界城市规模测试平台）[06:35] 🧩 Teaching LLMs a Low-Resource Language: Enhancing Code Completion in Pharo（教会大语言模型一种低资源语言：提升Pharo语言中的代码补全能力）【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

【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：[00:30] 🎮 AlayaWorld: Long-Horizon and Playable Video World Generation（AlayaWorld：长视界且可游玩的视频世界生成）[01:31] 🤖 RynnWorld-4D: 4D Embodied World Models for Robotic Manipulation（RynnWorld-4D：面向机器人操作的4D具身世界模型）[02:30] 🔍 Hierarchical Sparse Attention Done Right: Toward Infinite Context Modeling（分层稀疏注意力机制的正确实现：迈向无限上下文建模）[03:36] 👁 Vision as Unified Multimodal Generation（视觉作为统一多模态生成）[04:25] 🔦 Light-Omni: Reflex over Reasoning in Agentic Video Understanding with Long-Term Memory（光之全能：基于长期记忆的智能体视频理解中反思优于推理）[05:25] 🎥 Parallelized Autoregressive Decoding for Omni-Modal Dense Video Captioning（并行化自回归解码用于全模态密集视频字幕生成）[06:14] 🧠 Gemma 4 Technical Report（Gemma 4 技术报告）[07:11] ⚡ SkillOpt-Lite: Better and Faster Agent Self-evolution via One Line of Vibe（SkillOpt-Lite：通过一行“氛围”实现更优更快的智能体自我进化）[08:07] ⚡ DSpark: Confidence-Scheduled Speculative Decoding with Semi-Autoregressive Generation（DSpark：基于置信度调度的半自回归推测解码）[09:03] 🧠 MentalThink: Shaping Thoughts in Mental SVG World（MentalThink：在思维SVG世界中塑造思想）[09:57] 🤖 RynnWorld-Teleop: An Action-Conditioned World Model for Digital Teleoperation（RynnWorld-Teleop：一种用于数字遥操作的动作条件世界模型）[10:58] 🎯 TurnOPD: Making On-Policy Distillation Turn-Aware for Efficient Long-Horizon Agent Training（TurnOPD：使在线知识蒸馏具备回合感知能力以实现高效的长时程智能体训练）[11:58] 🎨 CanvasAgent: Enabling Complex Image Creation and Editing via Visual Tool Orchestration（CanvasAgent：通过可视化工具编排实现复杂图像创建与编辑）[13:05] 🧪 TREK: Distill to Explore, Reinforce to Refine（TREK：通过蒸馏进行探索，通过强化进行精炼）[13:51] 🖼 PointDiT: Pixel-Space Diffusion for Monocular Geometry Estimation（PointDiT：面向单目几何估计的像素空间扩散模型）【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

【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 14 篇论文如下：[00:32] 🤖 UI-MOPD: Multi-Platform On-Policy Distillation for Continual GUI Agent Learning（UI-MOPD：面向持续GUI智能体学习的多平台在线策略蒸馏）[01:30] 💡 ResearchStudio-Idea: An Evidence-Grounded Research-Ideation Skill Suite from ML Conference Outcomes（ResearchStudio-Idea：基于证据的科研构思技能套件——来自机器学习会议成果）[02:21] 🎨 PixWorld: Unifying 3D Scene Generation and Reconstruction in Pixel Space（PixWorld：在像素空间中统一3D场景生成与重建）[03:11] 🧩 OmniOpt: Taxonomy, Geometry, and Benchmarking of Modern Optimizers（OmniOpt：现代优化器的分类、几何结构与基准测试）[04:04] 🤖 GigaWorld-1: A Roadmap to Build World Models for Robot Policy Evaluation（GigaWorld-1：构建用于机器人策略评估的世界模型路线图）[04:55] 🧩 Vision Pretraining for Dense Spatial Perception（面向密集空间感知的视觉预训练）[05:54] 🤖 EVA-Client: A Unified Data Collection, Inference, and Deployment Framework for Embodied Policies on Real Robots（EVA-Client：面向实体机器人上的具身策略的统一数据收集、推理与部署框架）[06:47] 🎥 Wan-Streamer v0.2: Higher Resolution, Same Latency（Wan-Streamer v0.2：更高分辨率，相同延迟）[07:48] 🤖 InternVLA-A1.5: Unifying Understanding, Latent Foresight, and Action for Compositional Generalization（InternVLA-A1.5：统一理解、潜在预知与动作以实现组合泛化）[08:46] 🔍 Do All Visual Tokens Matter Equally? Object-Evidence Preserving Token Merging for Vision-Language Retrieval（所有视觉标记都同等重要吗？面向视觉-语言检索的保留对象证据的标记合并方法）[09:59] 🧠 KVpop -- Key-Value Cache Compression with Predictive Online Pruning（KVpop——基于预测性在线剪枝的键值缓存压缩）[10:50] 🧠 dOPSD: On-Policy Self-Distillation for Diffusion Language Models（dOPSD：扩散语言模型的在线自蒸馏方法）[11:41] 🎨 Perceptual Flow Matching for Few-Step Generative Modeling（感知流匹配：用于少步生成建模）[12:37] 🔬 Multi-Turn Agentic Scientific Literature Search via Workflow Induction（多轮交互式科学文献搜索：基于工作流归纳的方法）【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

【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 9 篇论文如下：[00:35] 🎯 The Mirage of Optimizing Training Policies: Monotonic Inference Policies as the Real Objective for LLM Reinforcement Learning（优化训练策略的幻影：单调推理策略作为大语言模型强化学习的真正目标）[01:34] 🛠 VLA-Corrector: Lightweight Detect-and-Correct Inference for Adaptive Action Horizon（VLA-Corrector：一种用于自适应动作视界的轻量级检测与校正推理框架）[02:35] 🤖 Embodied.cpp: A Portable Inference Runtime of Embodied AI Models on Heterogeneous Robots（Embodied.cpp：面向异构机器人的具身AI模型便携推理运行时）[03:43] 🔢 OrbitQuant: Data-Agnostic Quantization for Image and Video Diffusion Transformers（OrbitQuant：面向图像和视频扩散Transformer的数据无关量化方法）[04:38] 📊 DataComp-VLM: Improved Open Datasets for Vision-Language Models（DataComp-VLM：改进的视觉语言模型开放数据集）[05:36] 🛡 Securing the AI Agent: A Unified Framework for Multi-Layer Agent Red Teaming（保护AI智能体：一种多层次智能体红队测试的统一框架）[06:36] ☁ Interpretation-Oriented Cloud Removal via Observation-Anchored Residual Flow with Geo-Contextual Alignment（面向解译的云去除：基于观测锚定残差流与地理上下文对齐的方法）[07:38] 📄 MultAttnAttrib: Training-Free Multimodal Attribution in Long Document Question Answering（多注意力归因：长文档问答中的无训练多模态归因）[08:25] 🔗 AGE: Adaptive-masking for Graph Embedding in Graph Retrieval-Augmented Generation（AGE：面向图检索增强生成的自适应掩码图嵌入方法）【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

【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：[00:43] TOP1(🔥230) | 🌍 Orca: The World is in Your Mind（虎鲸：世界在你心中）[03:18] TOP2(🔥141) | 🤔 Agentic Abstention: Do Agents Know When to Stop Instead of Act?（智能体式弃权：智能体知道何时该停止而非行动吗？）[05:32] TOP3(🔥103) | 🧪 Dockerless: Environment-Free Program Verifier for Coding Agents（无Docker：面向编码智能体的无环境程序验证器）[07:58] TOP4(🔥93) | 🎭 DOPD: Dual On-policy Distillation（双在线策略蒸馏）[10:13] TOP5(🔥86) | 🧠 Scaling the Horizon, Not the Parameters: Reaching Trillion-Parameter Performance with a 35B Agent（扩展智能体视野而非参数规模：以35B智能体达到万亿参数级性能）【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