Can VLMs Actually See and Read? A Survey on Modality Collapse in Vision-Language Models.

Mong Yuan Sim, Wei Emma Zhang, Xiang Dai, Biaoyan Fang

Browse the full ACL paper archive.