0.5概率即逐字背诵，倒放视频测因果

重点关注

• AgentDoG 1.5: A Lightweight and Scalable Alignment Framework for AI Agent Safety and Security score 8
  入选 HF Daily Papers；HF 热度: 105 upvotes (+4)；关键词(4): lightweight, deployment, real-time, agentic
• minWM: A Full-Stack Open-Source Framework for Real-Time Interactive Video World Models score 10
  入选 HF Daily Papers；HF 热度: 44 upvotes (+4)；有代码实现；关键词(5): distillation, latency, real-time, fine-tuning, open-source
• YoCausal: How Far is Video Generation from World Model? A Causality Perspective score 10
  入选 HF Daily Papers；HF 热度: 37 upvotes (+4)；有代码实现；关键词(2): reasoning, synthetic data
• GenClaw: Code-Driven Agentic Image Generation score 10
  入选 HF Daily Papers；HF 热度: 30 upvotes (+4)；有代码实现；关键词(2): agentic, reasoning
• EarlyTom: Early Token Compression Completes Fast Video Understanding score 10
  入选 HF Daily Papers；HF 热度: 24 upvotes (+4)；有代码实现；关键词(4): compression, deployment, production, throughput
• How LoRA Remembers? A Parametric Memory Law for LLM Finetuning score 10
  入选 HF Daily Papers；HF 热度: 22 upvotes (+4)；有代码实现；关键词(1): finetuning
• OmniRetrieval: Unified Retrieval across Heterogeneous Knowledge Sources score 9
  入选 HF Daily Papers；HF 热度: 61 upvotes (+4)；有代码实现
• Native Audio-Visual Alignment for Generation score 10
  入选 HF Daily Papers；HF 热度: 22 upvotes (+4)；有代码实现；关键词(1): open-source
• LoMo: Local Modality Substitution for Deeper Vision-Language Fusion score 9
  入选 HF Daily Papers；HF 热度: 18 upvotes (+3)；有代码实现；关键词(5): lightweight, serving, reasoning, vision-language, data curation
• Xetrieval: Mechanistically Explaining Dense Retrieval score 9
  入选 HF Daily Papers；HF 热度: 15 upvotes (+3)；有代码实现；关键词(2): lightweight, reasoning

也值得关注

• NeuROK: Generative 4D Neural Object Kinematics score 4
  入选 HF Daily Papers；HF 热度: 4 upvotes (+1)
• AdaState: Self-Evolving Anchors for Streaming Video Generation score 5
  入选 HF Daily Papers；HF 热度: 5 upvotes (+2)
• CoHyDE: Iterative Co-Training of LLM Rewriter & Dense Encoder for Tool Retrieval score 5
  入选 HF Daily Papers；HF 热度: 2 upvotes (+1)；关键词(2): fine-tuning, DPO
• Beyond 3D VQAs: Injecting 3D Spatial Priors into Vision-Language Models for Enhanced Geometric Reasoning score 4
  入选 HF Daily Papers；关键词(3): fine-tuning, reasoning, vision-language
• VFEAgent: A Multimodal Agent Framework for End-to-End Automated Finite Element Analysis score 4
  机构: Peking University；关键词(2): reasoning, vision-language
• SURGENT: A Surgical Multi-Agent Assistance System Across the Perioperative Workflow score 4
  机构: Huawei；关键词(3): deployment, retrieval-augmented, reasoning
• MOOSE-Copilot: A Web-Based Interactive Assistant for Unified Exploratory and Fine-Grained Scientific Hypothesis Discovery score 4
  关键词(1): agentic；顶会接收: ACL