AI数学研究agent解开真实开放问题，世界模型扎堆涌现

重点关注

• Towards Autonomous Mathematics Research score 12
  机构: Google；入选 HF Daily Papers；HF 热度: 29 upvotes (+4)；关键词(6): scaling, inference, agent, tool use, reasoning
• Stroke3D: Lifting 2D strokes into rigged 3D model via latent diffusion models score 10
  入选 HF Daily Papers；HF 热度: 8 upvotes (+2)；关键词(4): DPO, alignment, preference, diffusion；顶会接收: ICLR
• Code2World: A GUI World Model via Renderable Code Generation score 9
  入选 HF Daily Papers；HF 热度: 184 upvotes (+4)；关键词(3): agents, code generation, vision-language
• Chain of Mindset: Reasoning with Adaptive Cognitive Modes score 9
  入选 HF Daily Papers；HF 热度: 69 upvotes (+4)；关键词(5): efficiency, agent, agentic, code generation, reasoning
• P1-VL: Bridging Visual Perception and Scientific Reasoning in Physics Olympiads score 9
  入选 HF Daily Papers；HF 热度: 56 upvotes (+4)；关键词(10): serving, inference, post-training, agent, agentic
• Agent World Model: Infinity Synthetic Environments for Agentic Reinforcement Learning score 9
  入选 HF Daily Papers；HF 热度: 47 upvotes (+4)；关键词(6): scaling, efficient, agent, agents, agentic
• Olaf-World: Orienting Latent Actions for Video World Modeling score 9
  入选 HF Daily Papers；HF 热度: 25 upvotes (+4)；关键词(3): scaling, efficient, alignment
• VLA-JEPA: Enhancing Vision-Language-Action Model with Latent World Model score 8
  入选 HF Daily Papers；HF 热度: 15 upvotes (+3)；关键词(3): fine-tuning, pretraining, vision-language
• Fine-T2I: An Open, Large-Scale, and Diverse Dataset for High-Quality T2I Fine-Tuning score 8
  入选 HF Daily Papers；HF 热度: 13 upvotes (+3)；关键词(6): fine-tuning, alignment, pretraining, diffusion, text-to-image
• VideoWorld 2: Learning Transferable Knowledge from Real-world Videos score 8
  入选 HF Daily Papers；HF 热度: 13 upvotes (+3)；关键词(4): diffusion, agents, reasoning, robotics

也值得关注

• Monocular Normal Estimation via Shading Sequence Estimation score 4
  关键词(1): benchmark；顶会接收: ICLR
• AnyTouch 2: General Optical Tactile Representation Learning For Dynamic Tactile Perception score 3
  顶会接收: ICLR
• Human Control Is the Anchor, Not the Answer: Early Divergence of Oversight in Agentic AI Communities score 2
  关键词(4): deployment, agent, agentic, guardrails
• Measuring Privacy Risks and Tradeoffs in Financial Synthetic Data Generation score 2
  关键词(2): diffusion, synthetic data
• Positive-Unlabelled Active Learning to Curate a Dataset for Orca Resident Interpretation score 2
  关键词(4): efficiency, transformer, audio, search
• Stabilizing Physics-Informed Consistency Models via Structure-Preserving Training score 2
  关键词(5): fast, inference, fine-tuning, diffusion, cost
• Statistical Roughness-Informed Machine Unlearning score 2
  关键词(3): inference, alignment, evaluation
• Reward Modeling for Reinforcement Learning-Based LLM Reasoning: Design, Challenges, and Evaluation score 2
  关键词(7): scaling, efficient, inference, fine-tuning, alignment
• Empowering Contrastive Federated Sequential Recommendation with LLMs score 2
  关键词(2): preference, recommendation
• How Far Can You Grow? Characterizing the Extrapolation Frontier of Graph Generative Models for Materials Science score 2
  关键词(3): scaling, benchmark, evaluation
• Clarifying Shampoo: Adapting Spectral Descent to Stochasticity and the Parameter Trajectory score 2
  关键词(2): efficient, efficiency
• Effective MoE-based LLM Compression by Exploiting Heterogeneous Inter-Group Experts Routing Frequency and Information Density score 2
  关键词(5): efficient, compression, deployment, post-training, MoE
• SnareNet: Flexible Repair Layers for Neural Networks with Hard Constraints score 2
  关键词(2): planning, safety
• GAFR-Net: A Graph Attention and Fuzzy-Rule Network for Interpretable Breast Cancer Image Classification score 2
  关键词(3): attention, reasoning, benchmark