1 paper
Souraja Kundu, Aditya Gupta, Joel Julin +3
Multimodal large language models (MLLMs) are good at video understanding and code generation, but it remains unclear whether they can recover the physical laws underlying videos, a…