Multimodal RewardBench: Holistic Evaluation of Reward Models for Vision Language Models

Open in new window