2Rethinking RAG in Long Videos: What to Retrieve and How to Use It?（长视频中重新思考RAG：要检索什么以及如何使用？）HF Paperspapers
3Smaller Models are Natural Explorers for Policy-Level Diversity in GRPO（较小的模型是GRPO策略级多样性自然探索者）HF Paperspapers
4Measuring Epistemic Resilience of LLMs Under Misleading Medical Context（测量在误导性医疗情境下大型语言模型的 epistemic 稳定性）HF Paperspapers
5RhymeFlow: Training-Free Acceleration for Video Generation with Asynchronous Denoising Flow Scheduling（RhymeFlow：无需训练的加速——基于异步去噪流调度的视频生成）HF Paperspapers
6MBench: A Comprehensive Benchmark on Memory Capability for Video World Models（MBench：视频世界模型的全面内存能力基准）HF Paperspapers
7RepFusion: Leveraging Multimodal Priors for Denoising in Representation Space（RepFusion：在表示空间中利用多模态先验进行去噪）HF Paperspapers
8RedAct: Redacting Agent Capability Traces for Procedural Skill Protection（红化剂：用于保护程序技能的抹除代理能力跟踪）HF Paperspapers
9OmniDirector: General Multi-Shot Camera Cloning without Cross-Paired Data（OmniDirector：通用多视角相机克隆无需跨配对数据）HF Paperspapers
12HarnessX: A Composable, Adaptive, and Evolvable Agent Harness Foundry（HarnessX：一个可组合、自适应和可进化的代理夹具发现库）HF Paperspapers
13Hy-Embodied-0.5-VLA: From Vision-Language-Action Models to a Real-World Robot Learning Stack（Hy-Embodied-0.5-VLA：从视觉语言行动模型到真实世界机器人学习栈）HF Paperspapers
14Dense Supervision, Sparse Updates: On the Sparsity and Geometry of On-Policy Distillation（密集监督，稀疏更新：关于在线策略蒸馏的稀疏性和几何结构）HF Paperspapers
15Avatar V: Scaling Video-Reference Avatar Video Generation（Avatar V：缩放视频参考avatar视频生成）HF Paperspapers
16ClinHallu: A Benchmark for Diagnosing Stage-Wise Hallucinations in Medical MLLM Reasoning（ClinHallu：医疗MLLM推理中阶段式幻觉诊断的标准数据集）HF Paperspapers
20OmniVideo-100K: A Dataset for Audio-Visual Reasoning through Structured Scripts and Evidence Chains（OmniVideo-100K：一种基于结构化脚本和证据链的音频-视觉推理数据集）HF Paperspapers
24From Chatbot to Digital Colleague: The Paradigm Shift Toward Persistent Autonomous AIHF Paperspapers