1 paper
Orr Zohar, Xiaohan Wang, Yann Dubois +9
Despite the rapid integration of video perception capabilities into Large Multimodal Models (LMMs), the underlying mechanisms driving their video understanding remain poorly unders…