Yangzhou Liu

Will Multimodal Models Be Dazzled by Multi-Image Visual Puzzles?