AI RESEARCH
PhysNote: Self-Knowledge Notes for Evolvable Physical Reasoning in Vision-Language Model
arXiv CS.AI
•
ArXi:2604.24443v1 Announce Type: new Vision-Language Models (VLMs) have nstrated strong performance on textbook-style physics problems, yet they frequently fail when confronted with dynamic real-world scenarios that require temporal consistency and causal reasoning across frames.