HuggingFace 每日AI论文速递

本期的 15 篇论文如下：[00:19] 🧠 Every Attention Matters: An Efficient Hybrid Architecture for Long-Context Reasoning（每一种注意力都重要：面向长上下文推理的高效混合架构）[00:59] ⚖ BAPO: Stabilizing Off-Policy Reinforcement Learning for LLMs via Balanced Policy Optimization with Adaptive Clipping（BAPO：通过自适应裁剪的平衡策略优化稳定LLM离策略强化学习）[01:40] 🧠 LoongRL:Reinforcement Learning for Advanced Reasoning over Long Contexts（LoongRL：面向长文本高级推理的强化学习方法）[02:18] 🌍 GigaBrain-0: A World Model-Powered Vision-Language-Action Model（GigaBrain-0：基于世界模型的通才视觉-语言-动作大模型）[02:49] 🔄 Language Models are Injective and Hence Invertible（语言模型是单射的，因此可逆）[03:25] 📹 VideoAgentTrek: Computer Use Pretraining from Unlabeled Videos（VideoAgentTrek：利用无标注视频预训练计算机操作智能体）[04:01] 📲 DaMo: Data Mixing Optimizer in Fine-tuning Multimodal LLMs for Mobile Phone Agents（DaMo：面向手机智能体的多模态大模型微调数据配比优化器）[04:55] 🚀 Unified Reinforcement and Imitation Learning for Vision-Language Models（统一强化与模仿学习的视觉-语言模型）[05:28] 🖼 Pico-Banana-400K: A Large-Scale Dataset for Text-Guided Image Editing（Pico-Banana-400K：面向文本引导图像编辑的大规模高质量数据集）[06:17] 📊 FinSight: Towards Real-World Financial Deep Research（FinSight：迈向真实场景的金融深度研究）[07:06] 🧠 Are they lovers or friends? Evaluating LLMs' Social Reasoning in English and Korean Dialogues（他们是恋人还是朋友？评估大语言模型在英韩对话中的社会推理能力）[07:43] 🌍 OmniNWM: Omniscient Driving Navigation World Models（OmniNWM：全景驾驶导航全知世界模型）[08:28] 🕳 Attention Sinks in Diffusion Language Models（扩散语言模型中的注意力沉陷现象）[09:04] 📄 olmOCR 2: Unit Test Rewards for Document OCR（olmOCR 2：基于单元测试奖励的文档OCR系统）[09:42] 🧠 KORE: Enhancing Knowledge Injection for Large Multimodal Models via Knowledge-Oriented Augmentations and Constraints（KORE：通过知识导向增强与约束为大模型持续注入知识）【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

本期的 14 篇论文如下：[00:19] 🧠 LightMem: Lightweight and Efficient Memory-Augmented Generation（LightMem：轻量高效的记忆增强生成框架）[00:55] 🌀 World-in-World: World Models in a Closed-Loop World（世界中的世界：闭环环境下的世界模型）[01:44] 🖼 UniGenBench++: A Unified Semantic Evaluation Benchmark for Text-to-Image Generation（UniGenBench++：面向文本到图像生成的统一语义评测基准）[02:29] 🧪 Chem-R: Learning to Reason as a Chemist（Chem-R：像化学家一样学习推理）[03:10] 🎬 MoGA: Mixture-of-Groups Attention for End-to-End Long Video Generation（MoGA：面向端到端长视频生成的分组混合注意力机制）[03:52] 🔍 Grasp Any Region: Towards Precise, Contextual Pixel Understanding for Multimodal LLMs（任意区域皆可掌握：面向多模态大模型的精准上下文像素级理解）[04:49] 🎬 IF-VidCap: Can Video Caption Models Follow Instructions?（IF-VidCap：视频字幕模型能听懂指令吗？）[05:35] 🚀 Every Step Evolves: Scaling Reinforcement Learning for Trillion-Scale Thinking Model（万亿参数思维模型的强化学习扩展之路）[06:21] 🎬 MT-Video-Bench: A Holistic Video Understanding Benchmark for Evaluating Multimodal LLMs in Multi-Turn Dialogues（MT-Video-Bench：面向多轮对话评估多模态大模型视频理解能力的综合基准）[07:12] 🧠 ssToken: Self-modulated and Semantic-aware Token Selection for LLM Fine-tuning（ssToken：面向大模型微调的自调制语义感知Token筛选方法）[07:43] 🎬 MUG-V 10B: High-efficiency Training Pipeline for Large Video Generation Models（MUG-V 10B：面向大视频生成模型的高效训练流水线）[08:18] 🎯 ProCLIP: Progressive Vision-Language Alignment via LLM-based Embedder（ProCLIP：基于大语言模型嵌入器的渐进式视觉-语言对齐方法）[09:29] 🎬 UltraGen: High-Resolution Video Generation with Hierarchical Attention（UltraGen：基于分层注意力的原生高分辨率视频生成）[10:15] 🔄 DSI-Bench: A Benchmark for Dynamic Spatial Intelligence（DSI-Bench：动态空间智能评测基准）【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

本期的 13 篇论文如下：[00:21] 🪞 PICABench: How Far Are We from Physically Realistic Image Editing?（PICABench：我们离物理真实的图像编辑还有多远？）[01:04] 🤖 DeepAnalyze: Agentic Large Language Models for Autonomous Data Science（DeepAnalyze：面向自主数据科学的智能体大模型）[01:50] 🗜 Glyph: Scaling Context Windows via Visual-Text Compression（Glyph：通过视觉-文本压缩扩展上下文窗口长度）[02:23] 🔍 Towards Mixed-Modal Retrieval for Universal Retrieval-Augmented Generation（面向通用检索增强生成的混合模态检索研究）[03:10] 🔗 When to Ensemble: Identifying Token-Level Points for Stable and Fast LLM Ensembling（何时集成：定位Token级位置实现稳定高效的大模型集成）[04:09] 🎯 Annotation-Efficient Universal Honesty Alignment（注释高效型通用诚实对齐）[04:49] 🖌 Uniworld-V2: Reinforce Image Editing with Diffusion Negative-aware Finetuning and MLLM Implicit Feedback（Uniworld-V2：借助扩散负感知微调与MLLM隐式反馈强化图像编辑）[05:46] 👁 RL makes MLLMs see better than SFT（强化学习让多模态大模型看得比监督微调更清楚）[06:33] 🚀 Visual Autoregressive Models Beat Diffusion Models on Inference Time Scaling（视觉自回归模型在推理时扩展上击败扩散模型）[07:09] 🎨 ConsistEdit: Highly Consistent and Precise Training-free Visual Editing（ConsistEdit：面向MM-DiT的高一致免训练视觉编辑）[07:56] 🔄 Deep Self-Evolving Reasoning（深度自演化推理）[08:22] 🧠 Beyond Pipelines: A Survey of the Paradigm Shift toward Model-Native Agentic AI（超越流水线：模型原生智能体AI范式转移综述）[09:07] 🔮 Chronos-2: From Univariate to Universal Forecasting（Chronos-2：从单变量到通用预测）【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

本期的 15 篇论文如下：[00:20] 🧠 A Theoretical Study on Bridging Internal Probability and Self-Consistency for LLM Reasoning（大模型推理中内部概率与自洽性桥接的理论研究）[01:04] 🌐 OmniVinci: Enhancing Architecture and Data for Omni-Modal Understanding LLM（OmniVinci：面向全模态理解大模型的架构与数据增强）[01:44] 🎬 Scaling Instruction-Based Video Editing with a High-Quality Synthetic Dataset（用百万级合成数据集放大指令式视频编辑）[02:28] ✂ NANO3D: A Training-Free Approach for Efficient 3D Editing Without Masks（NANO3D：无需训练与掩码的高效3D编辑新方法）[03:05] 🛰 Skyfall-GS: Synthesizing Immersive 3D Urban Scenes from Satellite Imagery（Skyfall-GS：仅凭卫星影像合成沉浸式3D城市场景）[03:41] ⚠ Emergent Misalignment via In-Context Learning: Narrow in-context examples can produce broadly misaligned LLMs（情境学习中的突发错位：狭窄示例可让大模型广泛失准）[04:18] 🧬 Latent Diffusion Model without Variational Autoencoder（无需变分自编码器的潜在扩散模型）[04:52] 📸 LightsOut: Diffusion-based Outpainting for Enhanced Lens Flare Removal（LightsOut：基于扩散的延展补全提升镜头眩光去除）[05:30] 🧠 MorphoBench: A Benchmark with Difficulty Adaptive to Model Reasoning（MorphoBench：随模型推理能力自适应难度的评测基准）[06:14] 🧠 A$^2$FM: An Adaptive Agent Foundation Model for Tool-Aware Hybrid Reasoning（A²FM：面向工具感知混合推理的自适应智能体基础模型）[06:56] 🗣 Language Models Model Language（语言模型即语言本身）[07:36] 🖼 BLIP3o-NEXT: Next Frontier of Native Image Generation（BLIP3o-NEXT：原生图像生成的下一个前沿）[08:30] 🌐 Paper2Web: Let's Make Your Paper Alive!（Paper2Web：让你的论文“活”起来！）[09:12] 🔬 Foundation Models for Scientific Discovery: From Paradigm Enhancement to Paradigm Transition（面向科学发现的基础模型：从范式增强到范式跃迁）[09:55] 🔍 Explore to Evolve: Scaling Evolved Aggregation Logic via Proactive Online Exploration for Deep Research Agents（探索以进化：通过主动在线探索扩展深度研究智能体的聚合逻辑）【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

本期的 5 篇论文如下：[00:40] TOP1(🔥154) | 🚀 QeRL: Beyond Efficiency -- Quantization-enhanced Reinforcement Learning for LLMs（QeRL：超越效率——面向大语言模型的量化增强强化学习）[02:19] TOP2(🔥138) | 🧠 Diffusion Transformers with Representation Autoencoders（基于表示自编码器的扩散Transformer）[04:54] TOP3(🔥134) | 🎯 Spatial Forcing: Implicit Spatial Representation Alignment for Vision-language-action Model（空间强迫：面向视觉-语言-动作模型的隐式空间表征对齐）[07:55] TOP4(🔥125) | 🖥 D2E: Scaling Vision-Action Pretraining on Desktop Data for Transfer to Embodied AI（D2E：利用桌面数据规模化视觉-动作预训练以迁移至具身智能）[10:30] TOP5(🔥110) | 📷 Thinking with Camera: A Unified Multimodal Model for Camera-Centric Understanding and Generation（基于相机的统一多模态理解与生成模型）【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

本期的 11 篇论文如下：[00:25] 👓 AI for Service: Proactive Assistance with AI Glasses（AI服务：AI眼镜的主动式协助）[01:06] 🎬 ImagerySearch: Adaptive Test-Time Search for Video Generation Beyond Semantic Dependency Constraints（ImagerySearch：面向超越语义依赖约束的自适应测试时搜索视频生成）[01:43] 🎯 LaSeR: Reinforcement Learning with Last-Token Self-Rewarding（LaSeR：基于末词元自奖励的强化学习）[02:33] 🧩 TokDrift: When LLM Speaks in Subwords but Code Speaks in Grammar（TokDrift：当大模型用子词而代码用语法时）[03:35] 🧠 Information Gain-based Policy Optimization: A Simple and Effective Approach for Multi-Turn LLM Agents（基于信息增益的策略优化：一种简单有效的多轮LLM智能体训练方法）[04:04] ⚡ Attention Is All You Need for KV Cache in Diffusion LLMs（扩散式大语言模型只需注意力即可搞定KV缓存）[04:45] 🤥 When Models Lie, We Learn: Multilingual Span-Level Hallucination Detection with PsiloQA（当模型撒谎时我们反而学到东西：用PsiloQA实现跨语言细粒度幻觉检测）[05:33] 📄 PaddleOCR-VL: Boosting Multilingual Document Parsing via a 0.9B Ultra-Compact Vision-Language Model（PaddleOCR-VL：以9亿参数超轻量多模态模型刷新多语言文档解析性能）[06:13] 🧠 VR-Thinker: Boosting Video Reward Models through Thinking-with-Image Reasoning（VR-Thinker：通过“边看边想”推理提升视频奖励模型）[06:52] 📐 MathCanvas: Intrinsic Visual Chain-of-Thought for Multimodal Mathematical Reasoning（MathCanvas：面向多模态数学推理的内生视觉思维链）[07:39] 🧠 COIG-Writer: A High-Quality Dataset for Chinese Creative Writing with Thought Processes（COIG-Writer：高质量中文创意写作数据集，附带思维过程）【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

本期的 15 篇论文如下：[00:21] 🎧 UniMoE-Audio: Unified Speech and Music Generation with Dynamic-Capacity MoE（UniMoE-Audio：基于动态容量MoE的统一语音与音乐生成模型）[00:57] 🔍 Attention Illuminates LLM Reasoning: The Preplan-and-Anchor Rhythm Enables Fine-Grained Policy Optimization（注意力照亮大模型推理：预规划-锚定节奏实现细粒度策略优化）[01:38] ⚡ FlashWorld: High-quality 3D Scene Generation within Seconds（FlashWorld：秒级高质量3D场景生成）[02:06] 🐝 Bee: A High-Quality Corpus and Full-Stack Suite to Unlock Advanced Fully Open MLLMs（Bee：高质量语料与全栈套件解锁完全开源多模态大模型）[02:37] 🗣 InteractiveOmni: A Unified Omni-modal Model for Audio-Visual Multi-turn Dialogue（InteractiveOmni：面向音视频多轮对话的统一全模态模型）[03:24] 🌍 PhysMaster: Mastering Physical Representation for Video Generation via Reinforcement Learning（PhysMaster：通过强化学习掌握视频生成的物理表征）[04:00] 🧪 LIBERO-Plus: In-depth Robustness Analysis of Vision-Language-Action Models（LIBERO-Plus：视觉-语言-动作模型鲁棒性深度剖析）[04:43] 🚗 CVD-STORM: Cross-View Video Diffusion with Spatial-Temporal Reconstruction Model for Autonomous Driving（CVD-STORM：面向自动驾驶的跨视角视频扩散时空重建模型）[05:21] 🔍 Generative Universal Verifier as Multimodal Meta-Reasoner（生成式通用验证器：多模态元推理的反思引擎）[06:07] ⚖ ParallelBench: Understanding the Trade-offs of Parallel Decoding in Diffusion LLMs（ParallelBench：探明扩散式大模型并行解码的取舍）[06:43] 🎞 Trace Anything: Representing Any Video in 4D via Trajectory Fields（任意视频4D轨迹场表示：一次前馈即可还原每像素连续时空路径）[07:27] 🌍 Reasoning in Space via Grounding in the World（基于世界锚定的空间推理）[07:54] 🧠 The Role of Computing Resources in Publishing Foundation Model Research（计算资源在基础模型研究发表中的角色）[08:28] ⚖ UniME-V2: MLLM-as-a-Judge for Universal Multimodal Embedding Learning（UniME-V2：用多模态大模型当裁判，打造通用多模态表征）[09:05] 🤖 InternVLA-M1: A Spatially Guided Vision-Language-Action Framework for Generalist Robot Policy（InternVLA-M1：面向通用机器人策略的空间引导视觉-语言-动作框架）【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

本期的 14 篇论文如下：[00:20] 🖼 Advancing End-to-End Pixel Space Generative Modeling via Self-supervised Pre-training（通过自监督预训练推进端到端像素空间生成建模）[00:53] 📚 DITING: A Multi-Agent Evaluation Framework for Benchmarking Web Novel Translation（DITING：面向网络小说翻译评测的多智能体基准框架）[01:41] 🌐 Scaling Language-Centric Omnimodal Representation Learning（以语言为中心的跨模态表征扩展学习）[02:29] 🎯 Detect Anything via Next Point Prediction（通过下一点预测检测万物）[03:02] ⚡ FlashVSR: Towards Real-Time Diffusion-Based Streaming Video Super-Resolution（FlashVSR：迈向实时扩散式流媒体视频超分辨率）[03:40] 🎯 Temporal Alignment Guidance: On-Manifold Sampling in Diffusion Models（时间对齐引导：扩散模型中的流形采样）[04:16] 🧠 Dr.LLM: Dynamic Layer Routing in LLMs（Dr.LLM：大模型中的动态层级路由）[05:03] 🎯 Spatial Forcing: Implicit Spatial Representation Alignment for Vision-language-action Model（空间强迫：面向视觉-语言-动作模型的隐式空间表征对齐）[05:50] 🤖 ERA: Transforming VLMs into Embodied Agents via Embodied Prior Learning and Online Reinforcement Learning（ERA：借助具身先验学习与在线强化学习将视觉-语言模型转化为具身智能体）[06:35] 🤖 Robot Learning: A Tutorial（机器人学习教程：从强化学习到多任务通用模型）[07:27] 🔄 SRUM: Fine-Grained Self-Rewarding for Unified Multimodal Models（SRUM：面向统一多模态模型的细粒度自奖励机制）[08:01] 🧠 Boundary-Guided Policy Optimization for Memory-efficient RL of Diffusion Large Language Models（面向扩散大语言模型的边界引导策略优化：内存高效的强化学习）[09:06] 🖼 UniFusion: Vision-Language Model as Unified Encoder in Image Generation（UniFusion：将视觉-语言模型统一作为图像生成的编码器）[09:43] 🧠 Memory as Action: Autonomous Context Curation for Long-Horizon Agentic Tasks（记忆即行动：面向长程智能体任务的自主上下文策展）【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

本期的 15 篇论文如下：[00:23] 🚀 QeRL: Beyond Efficiency -- Quantization-enhanced Reinforcement Learning for LLMs（QeRL：超越效率——面向大语言模型的量化增强强化学习）[01:22] 🧠 Diffusion Transformers with Representation Autoencoders（基于表示自编码器的扩散Transformer）[02:12] 🎬 OmniVideoBench: Towards Audio-Visual Understanding Evaluation for Omni MLLMs（OmniVideoBench：面向全向多模态大模型的音视频协同理解评测基准）[02:41] 🔄 Latent Refinement Decoding: Enhancing Diffusion-Based Language Models by Refining Belief States（潜变量精化解码：通过精化信念状态增强基于扩散的语言模型）[03:18] 🌊 RLFR: Extending Reinforcement Learning for LLMs with Flow Environment（RLFR：基于潜流环境扩展大模型强化学习）[04:11] 🔍 Spotlight on Token Perception for Multimodal Reinforcement Learning（多模态强化学习中token感知的光束聚焦）[04:50] 🎬 AVoCaDO: An Audiovisual Video Captioner Driven by Temporal Orchestration（AVoCaDO：面向时序编排的音视频联合字幕生成器）[05:25] 🌐 DiT360: High-Fidelity Panoramic Image Generation via Hybrid Training（DiT360：混合训练视角与全景数据的高保真全景图像生成）[05:56] 🧠 Demystifying Reinforcement Learning in Agentic Reasoning（揭开强化学习在智能体推理中的神秘面纱）[06:51] 🧮 Making Mathematical Reasoning Adaptive（让数学推理具备自适应性）[07:26] 🛡 Building a Foundational Guardrail for General Agentic Systems via Synthetic Data（面向通用智能体的基础护栏：基于合成数据的预执行安全框架）[08:05] 🧠 ACADREASON: Exploring the Limits of Reasoning Models with Academic Research Problems（ACADREASON：用学术研究问题探索推理模型的极限）[08:43] 🎨 InternSVG: Towards Unified SVG Tasks with Multimodal Large Language Models（InternSVG：用多模态大模型统一搞定SVG理解、编辑与生成）[09:23] 🧾 FinAuditing: A Financial Taxonomy-Structured Multi-Document Benchmark for Evaluating LLMs（FinAuditing：面向LLM评估的财务分类多文档基准）[10:09] 🧠 GIR-Bench: Versatile Benchmark for Generating Images with Reasoning（GIR-Bench：面向推理图像生成的多功能基准）【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

本期的 14 篇论文如下：[00:20] 🖥 D2E: Scaling Vision-Action Pretraining on Desktop Data for Transfer to Embodied AI（D2E：利用桌面数据规模化视觉-动作预训练以迁移至具身智能）[01:13] 📷 Thinking with Camera: A Unified Multimodal Model for Camera-Centric Understanding and Generation（基于相机的统一多模态理解与生成模型）[01:56] 🎨 TAG:Tangential Amplifying Guidance for Hallucination-Resistant Diffusion Sampling（TAG：抑制幻觉的扩散采样切向放大引导）[02:31] 🧠 Multimodal Prompt Optimization: Why Not Leverage Multiple Modalities for MLLMs（多模态提示优化：为何不为多模态大模型释放全模态潜能）[03:05] 🚀 AutoPR: Let's Automate Your Academic Promotion!（AutoPR：让学术晋升一键自动化！）[03:39] 🧭 R-Horizon: How Far Can Your Large Reasoning Model Really Go in Breadth and Depth?（R-HORIZON：你的大推理模型在广度与深度上究竟能走多远？）[04:14] 🚀 Webscale-RL: Automated Data Pipeline for Scaling RL Data to Pretraining Levels（Webscale-RL：把强化学习数据扩展到预训练体量的自动化流水线）[04:56] 🛰 SpaceVista: All-Scale Visual Spatial Reasoning from mm to km（SpaceVista：毫米到千米全尺度视觉空间推理）[05:37] 🎥 StreamingVLM: Real-Time Understanding for Infinite Video Streams（StreamingVLM：面向无限视频流的实时理解框架）[06:19] 🌐 KORMo: Korean Open Reasoning Model for Everyone（KORMo：人人可用的韩语开放推理模型）[06:42] ♻ Don't Waste Mistakes: Leveraging Negative RL-Groups via Confidence Reweighting（别浪费错误：通过置信度加权利用负RL组）[07:25] 🧠 Bridging Reasoning to Learning: Unmasking Illusions using Complexity Out of Distribution Generalization（从推理到学习的桥梁：以复杂度分布外泛化揭穿幻觉）[08:16] ⚡ DISCO: Diversifying Sample Condensation for Efficient Model Evaluation（DISCO：以模型分歧为导向的样本浓缩加速评测）[08:56] 🚗 Progressive Gaussian Transformer with Anisotropy-aware Sampling for Open Vocabulary Occupancy Prediction（面向开放词汇占用预测的各向异性采样渐进高斯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

本期的 5 篇论文如下：[00:33] TOP1(🔥300) | 🧠 Less is More: Recursive Reasoning with Tiny Networks（小而精：用微型网络递归推理）[02:16] TOP2(🔥164) | 🌱 Agent Learning via Early Experience（基于早期经验的主体学习）[04:15] TOP3(🔥105) | 🧠 Apriel-1.5-15b-Thinker（Apriel-1.5-15B-Thinker：以小博大实现前沿多模态推理的15B开源模型）[06:17] TOP4(🔥97) | 🧠 MM-HELIX: Boosting Multimodal Long-Chain Reflective Reasoning with Holistic Platform and Adaptive Hybrid Policy Optimization（MM-HELIX：以整体平台与自适应混合策略优化激发多模态长链反思推理）[08:45] TOP5(🔥88) | 🎬 Paper2Video: Automatic Video Generation from Scientific Papers（论文自动生成学术演讲视频）【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

本期的 14 篇论文如下：[00:16] 🌱 Agent Learning via Early Experience（基于早期经验的主体学习）[00:50] 🧠 MM-HELIX: Boosting Multimodal Long-Chain Reflective Reasoning with Holistic Platform and Adaptive Hybrid Policy Optimization（MM-HELIX：以整体平台与自适应混合策略优化激发多模态长链反思推理）[01:42] 🧪 From What to Why: A Multi-Agent System for Evidence-based Chemical Reaction Condition Reasoning（从“是什么”到“为什么”：面向循证化学反应条件推理的多智能体系统）[02:19] 🎬 UniVideo: Unified Understanding, Generation, and Editing for Videos（UniVideo：统一理解、生成与编辑视频的多模态框架）[03:01] 🧠 When Thoughts Meet Facts: Reusable Reasoning for Long-Context LMs（当思想邂逅事实：面向长上下文语言模型的可复用推理）[03:43] 🧠 Meta-Awareness Enhances Reasoning Models: Self-Alignment Reinforcement Learning（元认知增强推理模型：自对齐强化学习）[04:25] 🧠 MemMamba: Rethinking Memory Patterns in State Space Model（MemMamba：重新思考状态空间模型中的记忆模式）[05:17] 🛡 The Alignment Waltz: Jointly Training Agents to Collaborate for Safety（对齐圆舞曲：联合训练智能体协同守护安全）[05:53] 🎯 Hybrid Reinforcement: When Reward Is Sparse, It's Better to Be Dense（混合强化：奖励稀疏时，密集信号更胜一筹）[06:40] 🧪 NewtonBench: Benchmarking Generalizable Scientific Law Discovery in LLM Agents（NewtonBench：评测大模型智能体在通用科学定律发现中的基准）[07:17] 🪚 DeepPrune: Parallel Scaling without Inter-trace Redundancy（DeepPrune：并行扩展中消除跨路径冗余的高效推理框架）[07:54] 🚀 Training-Free Group Relative Policy Optimization（免训练群组相对策略优化）[08:24] 🪄 ARTDECO: Towards Efficient and High-Fidelity On-the-Fly 3D Reconstruction with Structured Scene Representation（ARTDECO：面向高效高保真即时三维重建的结构化场景表征）[08:55] 🤥 LLMs Learn to Deceive Unintentionally: Emergent Misalignment in Dishonesty from Misaligned Samples to Biased Human-AI Interactions（大模型在欺骗性样本与偏见人机交互中意外学会欺骗：不诚实行为的新兴错位）【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

本期的 15 篇论文如下：[00:21] 🔄 Ming-UniVision: Joint Image Understanding and Generation with a Unified Continuous Tokenizer（Ming-UniVision：用统一连续视觉词表打通图像理解与生成）[00:59] 🧠 Cache-to-Cache: Direct Semantic Communication Between Large Language Models（缓存到缓存：大模型间的直接语义通信）[01:32] 🌀 Lumina-DiMOO: An Omni Diffusion Large Language Model for Multi-Modal Generation and Understanding（Lumina-DiMOO：面向多模态生成与理解的离散扩散大模型）[02:07] 🧠 SHANKS: Simultaneous Hearing and Thinking for Spoken Language Models（SHANKS：口语模型边听边想的同步推理框架）[03:06] 🤖 RLinf-VLA: A Unified and Efficient Framework for VLA+RL Training（RLinf-VLA：面向VLA模型强化学习训练的统一高效框架）[04:02] 🎬 MATRIX: Mask Track Alignment for Interaction-aware Video Generation（MATRIX：面向交互感知视频生成的掩码轨迹对齐）[04:51] 🎯 Vibe Checker: Aligning Code Evaluation with Human Preference（Vibe Checker：让代码评估对齐人类偏好）[05:44] 🤖 Multi-Agent Tool-Integrated Policy Optimization（多智能体工具集成策略优化）[06:24] 🧠 CALM Before the STORM: Unlocking Native Reasoning for Optimization Modeling（风暴前夜：解锁优化建模原生推理潜能的轻量化矫正框架）[06:59] ✂ OBS-Diff: Accurate Pruning For Diffusion Models in One-Shot（OBS-Diff：一次性精准剪枝扩散模型）[07:52] 🧠 Artificial Hippocampus Networks for Efficient Long-Context Modeling（面向高效长上下文建模的人工海马网络）[08:30] 🔍 Revisiting Long-context Modeling from Context Denoising Perspective（基于上下文降噪视角的长文本建模再审视）[09:11] 🧠 Pushing on Multilingual Reasoning Models with Language-Mixed Chain-of-Thought（推动多语言推理模型：语言混合思维链新范式）[09:51] 💥 Why Low-Precision Transformer Training Fails: An Analysis on Flash Attention（低精度Transformer训练为何失败：Flash Attention失效机理剖析）[10:37] ⚡ Native Hybrid Attention for Efficient Sequence Modeling（原生混合注意力高效序列建模）【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

本期的 15 篇论文如下：[00:24] 📊 TaTToo: Tool-Grounded Thinking PRM for Test-Time Scaling in Tabular Reasoning（TaTToo：面向表格推理测试时扩展的“工具落地思维”过程奖励模型）[00:57] 🔍 Fathom-DeepResearch: Unlocking Long Horizon Information Retrieval and Synthesis for SLMs（Fathom-DeepResearch：解锁小模型长程信息检索与综合的钥匙）[01:39] 🚀 Fast-dLLM v2: Efficient Block-Diffusion LLM（Fast-dLLM v2：高效的块扩散大语言模型）[02:30] 🧑 CoDA: Coding LM via Diffusion Adaptation（CoDA：基于扩散适配的轻量级代码生成模型）[03:01] 🧩 Scaling Code-Assisted Chain-of-Thoughts and Instructions for Model Reasoning（规模化代码辅助思维链与指令以增强模型推理）[03:52] ⚖ ASPO: Asymmetric Importance Sampling Policy Optimization（ASPO：非对称重要性采样策略优化）[04:34] 🔗 Mixing Mechanisms: How Language Models Retrieve Bound Entities In-Context（混合机制：语言模型如何在上下文中检索绑定实体）[05:15] 🧠 AInstein: Assessing the Feasibility of AI-Generated Approaches to Research Problems（AInstein：评估AI生成科研方案可行性的研究框架）[05:51] 🪂 Refusal Falls off a Cliff: How Safety Alignment Fails in Reasoning?（拒绝断崖：安全对齐在推理中为何崩塌）[06:35] 🌍 HoloScene: Simulation-Ready Interactive 3D Worlds from a Single Video（HoloScene：单视频生成可交互3D仿真世界）[07:22] ⚡ TensorBLEU: Vectorized GPU-based BLEU Score Implementation for Per-Sentence In-Training Evaluation（TensorBLEU：面向逐句训练评估的向量化GPU加速BLEU分数实现）[08:09] 🎯 Margin Adaptive DPO: Leveraging Reward Model for Granular Control in Preference Optimization（边缘自适应DPO：利用奖励模型实现偏好优化的粒度控制）[09:00] 🩺 Discrete Diffusion Models with MLLMs for Unified Medical Multimodal Generation（基于多模态大语言模型的离散扩散模型实现统一医学多模态生成）[09:46] 🧠 MixReasoning: Switching Modes to Think（混合推理：动态切换思考模式）[10:20] ⚡ LightCache: Memory-Efficient, Training-Free Acceleration for Video Generation（LightCache：面向视频生成的内存高效、无需训练的加速方法）【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

本期的 15 篇论文如下：[00:21] 🎬 Paper2Video: Automatic Video Generation from Scientific Papers（论文自动生成学术演讲视频）[00:55] 🎬 Video-LMM Post-Training: A Deep Dive into Video Reasoning with Large Multimodal Models（Video-LMM后训练：深入剖析大型多模态模型的视频推理）[01:38] 🎬 VChain: Chain-of-Visual-Thought for Reasoning in Video Generation（VChain：面向视频生成推理的视觉思维链）[02:14] 👻 Imperceptible Jailbreaking against Large Language Models（针对大语言模型的隐形越狱攻击）[02:56] 🌳 MITS: Enhanced Tree Search Reasoning for LLMs via Pointwise Mutual Information（MITS：基于点互信息的树搜索增强大模型推理）[03:30] 🧬 Hybrid Architectures for Language Models: Systematic Analysis and Design Insights（语言模型混合架构：系统剖析与设计洞见）[04:07] 📊 Factuality Matters: When Image Generation and Editing Meet Structured Visuals（事实至关重要：当图像生成与编辑遇上结构化视觉）[04:59] 🔄 Reactive Transformer (RxT) -- Stateful Real-Time Processing for Event-Driven Reactive Language Models（反应式Transformer：事件驱动的实时有状态对话模型）[05:55] ⚖ Judging with Confidence: Calibrating Autoraters to Preference Distributions（置信评判：将自动评分器校准到偏好分布）[06:44] 🎯 Reinforce-Ada: An Adaptive Sampling Framework for Reinforce-Style LLM Training（Reinforce-Ada：面向Reinforce风格LLM训练的自适应采样框架）[07:27] 📏 Optimal Scaling Needs Optimal Norm（最优扩放需要最优范数）[07:51] 🔬 Code4MeV2: a Research-oriented Code-completion Platform（Code4MeV2：面向研究的代码补全平台）[08:31] 🪞 Self-Reflective Generation at Test Time（测试时自反思生成）[09:15] 🔄 SwiReasoning: Switch-Thinking in Latent and Explicit for Pareto-Superior Reasoning LLMs（SwiReasoning：在显式与潜空间之间切换思维，实现帕累托更优的推理大模型）[10:00] 👀 Watch and Learn: Learning to Use Computers from Online Videos（观看与学习：从在线视频中学习使用计算机）【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

本期的 15 篇论文如下：[00:28] 🧠 Apriel-1.5-15b-Thinker（Apriel-1.5-15B-Thinker：以小博大实现前沿多模态推理的15B开源模型）[01:04] 🚀 Efficient Multi-modal Large Language Models via Progressive Consistency Distillation（基于渐进一致性蒸馏的高效多模态大模型）[01:42] 🧩 Compose Your Policies! Improving Diffusion-based or Flow-based Robot Policies via Test-time Distribution-level Composition（组合式策略！利用测试时段分布级组合提升基于扩散或流的机器人策略性能）[02:19] 🪞 Self-Improvement in Multimodal Large Language Models: A Survey（多模态大语言模型自我提升综述）[02:59] 🧬 Your Agent May Misevolve: Emergent Risks in Self-evolving LLM Agents（你的智能体可能误入歧途：自演化大模型智能体中的涌现风险）[03:38] 📊 CoDA: Agentic Systems for Collaborative Data Visualization（CoDA：面向协同数据可视化的智能体系统）[04:21] 🧐 SurveyBench: How Well Can LLM(-Agents) Write Academic Surveys?（SurveyBench：大模型（智能体）写学术综述能有多靠谱？）[05:06] 🔧 REPAIR: Robust Editing via Progressive Adaptive Intervention and Reintegration（REPAIR：渐进式自适应干预与再融合的鲁棒编辑框架）[05:53] 🔍 OrtSAE: Orthogonal Sparse Autoencoders Uncover Atomic Features（OrtSAE：正交稀疏自编码器揭示原子级特征）[06:38] 🔍 FocusAgent: Simple Yet Effective Ways of Trimming the Large Context of Web Agents（FocusAgent：轻量级检索器为网页智能体精简冗长上下文的简易高效方案）[07:14] 🎯 Improving GUI Grounding with Explicit Position-to-Coordinate Mapping（基于显式位置-坐标映射的GUI定位改进方法）[08:05] 📏 LSPO: Length-aware Dynamic Sampling for Policy Optimization in LLM Reasoning（LSPO：面向大模型推理的基于长度感知的动态采样策略优化）[08:45] 🤖 WAInjectBench: Benchmarking Prompt Injection Detections for Web Agents（WAInjectBench：面向网页智能体的提示注入攻防基准评测）[09:19] 🍱 Free Lunch Alignment of Text-to-Image Diffusion Models without Preference Image Pairs（无需配对偏好图像即可免费对齐文本到图像扩散模型）[09:54] 🎯 LEAML: Label-Efficient Adaptation to Out-of-Distribution Visual Tasks for Multimodal Large Language Models（LEAML：面向多模态大模型的标签高效分布外视觉任务适配）【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

本期的 5 篇论文如下：[00:43] TOP1(🔥323) | 🐣 The Dragon Hatchling: The Missing Link between the Transformer and Models of the Brain（幼龙破壳： Transformer 与大脑模型之间缺失的环节）[02:38] TOP2(🔥167) | 🎬 LongLive: Real-time Interactive Long Video Generation（LongLive：实时交互式长视频生成框架）[05:04] TOP3(🔥150) | 🔥 MCPMark: A Benchmark for Stress-Testing Realistic and Comprehensive MCP Use（MCPMark：面向真实且全面的MCP应用场景的压力测试基准）[07:24] TOP4(🔥124) | 🧠 EPO: Entropy-regularized Policy Optimization for LLM Agents Reinforcement Learning（EPO：面向LLM智能体强化学习的熵正则策略优化）[09:18] TOP5(🔥122) | 🎮 Vision-Zero: Scalable VLM Self-Improvement via Strategic Gamified Self-Play（Vision-Zero：基于策略化博弈自对弈的可扩展视觉语言模型自我提升）【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

本期的 15 篇论文如下：[00:22] 🗜 LongCodeZip: Compress Long Context for Code Language Models（LongCodeZip：面向代码大模型的长上下文压缩方法）[00:56] 🎬 Self-Forcing++: Towards Minute-Scale High-Quality Video Generation（自增强++：迈向分钟级高质量视频生成）[01:38] 🧠 ExGRPO: Learning to Reason from Experience（基于经验的群体相对策略优化：让大模型学会从经验中推理）[02:32] 🥷 StealthAttack: Robust 3D Gaussian Splatting Poisoning via Density-Guided Illusions（隐身投毒：基于密度引导幻觉的鲁棒3D高斯溅射攻击）[03:32] 🎛 Interactive Training: Feedback-Driven Neural Network Optimization（交互式训练：反馈驱动的神经网络优化）[04:24] 📈 StockBench: Can LLM Agents Trade Stocks Profitably In Real-world Markets?（StockBench：大模型智能体能否在真实股市中稳定盈利？）[05:07] 🔍 VOGUE: Guiding Exploration with Visual Uncertainty Improves Multimodal Reasoning（VOGUE：用视觉不确定性引导探索，提升多模态推理）[05:44] 🪓 The Rogue Scalpel: Activation Steering Compromises LLM Safety（失控的手术刀：激活向量操控竟瓦解大模型安全锁）[06:21] 🔍 CLUE: Non-parametric Verification from Experience via Hidden-State Clustering（CLUE：基于隐状态聚类的非参数经验验证）[07:09] 🔍 ModernVBERT: Towards Smaller Visual Document Retrievers（ModernVBERT：打造更轻量的视觉文档检索器）[07:54] 🗺 RewardMap: Tackling Sparse Rewards in Fine-grained Visual Reasoning via Multi-Stage Reinforcement Learning（RewardMap：通过多阶段强化学习解决细粒度视觉推理中的稀疏奖励问题）[08:37] 🚀 F2LLM Technical Report: Matching SOTA Embedding Performance with 6 Million Open-Source Data（F2LLM技术报告：仅用600万开源数据即可达到SOTA嵌入性能）[09:13] 🧠 RLP: Reinforcement as a Pretraining Objective（RLP：将强化学习作为预训练目标）[09:45] 🖱 DragFlow: Unleashing DiT Priors with Region Based Supervision for Drag Editing（DragFlow：借助区域监督释放DiT先验，实现拖拽式编辑）[10:19] 🚀 The Unreasonable Effectiveness of Scaling Agents for Computer Use（扩展计算机使用代理的规模带来的不合理有效性）【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

本期的 15 篇论文如下：[00:19] 🧠 DeepSearch: Overcome the Bottleneck of Reinforcement Learning with Verifiable Rewards via Monte Carlo Tree Search（DeepSearch：以蒙特卡洛树搜索破解强化学习可验证奖励瓶颈）[01:20] 🤖 GEM: A Gym for Agentic LLMs（GEM：面向智能体大模型的开放训练场）[01:57] 🧠 VLA-RFT: Vision-Language-Action Reinforcement Fine-tuning with Verified Rewards in World Simulators（VLA-RFT：基于世界模拟器与验证奖励的视觉-语言-动作强化微调）[02:36] 🎒 Knapsack RL: Unlocking Exploration of LLMs via Optimizing Budget Allocation（背包强化学习：通过优化预算分配解锁大模型探索潜能）[03:06] 🎬 Code2Video: A Code-centric Paradigm for Educational Video Generation（Code2Video：面向教育视频生成的代码中心范式）[03:41] ⚙ PIPer: On-Device Environment Setup via Online Reinforcement Learning（PIPer：基于在线强化学习的设备端环境自动配置）[04:11] 🗜 ACON: Optimizing Context Compression for Long-horizon LLM Agents（ACON：面向长程LLM智能体的上下文压缩优化）[04:52] 🔍 Why Can't Transformers Learn Multiplication? Reverse-Engineering Reveals Long-Range Dependency Pitfalls（为何Transformer学不会乘法？逆向工程揭示长程依赖陷阱）[05:22] ⚖ BiasFreeBench: a Benchmark for Mitigating Bias in Large Language Model Responses（BiasFreeBench：面向大语言模型去偏响应评测的统一基准）[06:01] ⚡ Flash-Searcher: Fast and Effective Web Agents via DAG-Based Parallel Execution（Flash-Searcher：基于DAG并行执行的极速高效网络智能体）[06:42] 🚀 BroRL: Scaling Reinforcement Learning via Broadened Exploration（BroRL：通过拓宽探索规模来扩展强化学习）[07:25] 📊 Beyond Log Likelihood: Probability-Based Objectives for Supervised Fine-Tuning across the Model Capability Continuum（超越对数似然：面向模型能力连续谱的监督微调概率目标）[08:02] 🎯 On Predictability of Reinforcement Learning Dynamics for Large Language Models（论大型语言模型强化学习动力学的可预测性）[08:31] 🖥 GUI-KV: Efficient GUI Agents via KV Cache with Spatio-Temporal Awareness（GUI-KV：面向具备时空感知的高效GUI智能体的KV缓存方案）[09:17] 🧠 Training Vision-Language Process Reward Models for Test-Time Scaling in Multimodal Reasoning: Key Insights and Lessons Learned（训练视觉-语言过程奖励模型以实现多模态推理测试时扩展：关键洞见与经验总结）【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

本期的 10 篇论文如下：[00:29] TOP1(🔥640) | 🤝 Sharing is Caring: Efficient LM Post-Training with Collective RL Experience Sharing（共享即关爱：基于集体RL经验共享的高效大模型后训练）[02:49] TOP2(🔥341) | 🔒 A.S.E: A Repository-Level Benchmark for Evaluating Security in AI-Generated Code（A.S.E：一个用于评估AI生成代码安全的仓库级基准）[04:59] TOP3(🔥218) | 🤖 VLA-Adapter: An Effective Paradigm for Tiny-Scale Vision-Language-Action Model（VLA-Adapter：面向小型视觉-语言-动作模型的有效范式）[07:07] TOP4(🔥212) | 🤖 The Landscape of Agentic Reinforcement Learning for LLMs: A Survey（面向大语言模型的智能体强化学习全景：一项综述）[09:17] TOP5(🔥207) | 🤔 Drivel-ology: Challenging LLMs with Interpreting Nonsense with Depth（废话学：用深度解读无意义内容挑战大型语言模型）[11:19] TOP6(🔥183) | 🤔 Why Language Models Hallucinate（语言模型为何产生幻觉）[13:06] TOP7(🔥174) | 🧠 A Survey of Reinforcement Learning for Large Reasoning Models（大型推理模型的强化学习综述）[15:32] TOP8(🔥160) | 🎬 LongLive: Real-time Interactive Long Video Generation（LongLive：实时交互式长视频生成框架）[18:13] TOP9(🔥145) | 💡 Reverse-Engineered Reasoning for Open-Ended Generation（面向开放式生成的逆向工程推理）[20:27] TOP10(🔥140) | 🤖 A Survey of Scientific Large Language Models: From Data Foundations to Agent Frontiers（科学大型语言模型综述：从数据基础到智能体前沿）【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