Logic Unseen: Revealing the Logical Blindspots of Vision-Language Models

Yuchen Zhou; Jiayu Tang; Shuo Yang; Xiaoyan Xiao; Yuqin Dai; Wenhao Yang; Chao Gou; Xiaobo Xia

AAAI 2026technical0 citations

Logic Unseen: Revealing the Logical Blindspots of Vision-Language Models

Yuchen Zhou, Jiayu Tang, Shuo Yang, Xiaoyan Xiao, Yuqin Dai, Wenhao Yang, Chao Gou, Xiaobo Xia

Abstract

Vision-Language Models (VLMs), exemplified by CLIP, have emerged as foundational for multimodal intelligence. However, their capacity for logical understanding remains significantly underexplored, resulting in critical **

BibTeX

@inproceedings{aaai2026_logicunseenrevea,
  title = {Logic Unseen: Revealing the Logical Blindspots of Vision-Language Models},
  author = {Yuchen Zhou and Jiayu Tang and Shuo Yang and Xiaoyan Xiao and Yuqin Dai and Wenhao Yang and Chao Gou and Xiaobo Xia and Tat-Seng Chua},
  booktitle = {AAAI 2026},
  year = {2026}
}