AutoHallusion: Automatic Generation of Hallucination Benchmarks for Vision-Language Models

Benchmark Model Rank Results
visual-question-answering-vqa-on-5GPT-4V#1Overall Accuracy: 66.0