1 paper
Tiange Luo, Ang Cao, Gunhee Lee +2
Despite recent advances in Vision-Language Models (VLMs), they may over-rely on visual language priors existing in their training data rather than true visual reasoning. To investi…