Seeing Is Believing? A Benchmark for Multimodal Large Language Models on Visual Illusions and Anomalies

VLMLRMRe-assign community