paper-with-me

Papers

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

2025-02-20 · Michihiro Yasunaga, Luke Zettlemoyer, Marjan Ghazvininejad

Reward models play an essential role in training vision-language models (VLMs) by assessing output quality to enable aligning with human preferences. Despite their importance, the research community lacks comprehensive open benchmarks for evaluating multimodal reward models in VLMs. To address this gap, we introduce Multimodal RewardBench, an expert-annotated benchmark covering six domains: general correctness, preference, knowledge, reasoning, safety, and visual question-answering. Our dataset comprises 5,211 annotated (prompt, chosen response, rejected response) triplets collected from various VLMs. In evaluating a range of VLM judges, we find that even the top-performing models, Gemini 1.5 Pro and Claude 3.5 Sonnet, achieve only 72% overall accuracy. Notably, most models struggle in the reasoning and safety domains. These findings suggest that Multimodal RewardBench offers a challenging testbed for advancing reward model development across multiple domains. We release the benchmark at https://github.com/facebookresearch/multimodal_rewardbench.

📄 PDF Abstract BibTeX arXiv:2502.14191

Code (1)

facebookresearch/multimodal_rewardbench 공식 구현

Tasks

Question AnsweringVisual Question Answering

Similar Papers 제목 키워드 기반

VL-RewardBench: A Challenging Benchmark for Vision-Language Generative Reward Models

2025-01-01 · CVPR 2025 1 · Lei LI, Yuancheng Wei, Zhihui Xie, Xuqing Yang 외

Vision-language generative reward models (VL-GenRMs) play a crucial role in aligning and evaluating multimodal AI systems, yet their own evaluation remains under-explored. Current assessment methods primarily rely on…

Hallucination

VLRewardBench: A Challenging Benchmark for Vision-Language Generative Reward Models

2024-11-26 · Lei LI, Yuancheng Wei, Zhihui Xie, Xuqing Yang 외

Vision-language generative reward models (VL-GenRMs) play a crucial role in aligning and evaluating multimodal AI systems, yet their own evaluation remains under-explored. Current assessment methods primarily rely on AI-…

Hallucination

Skywork-VL Reward: An Effective Reward Model for Multimodal Understanding and Reasoning

2025-05-12 · Xiaokun Wang, Chris, Jiangbo Pei, Wei Shen 외

We propose Skywork-VL Reward, a multimodal reward model that provides reward signals for both multimodal understanding and reasoning tasks. Our technical approach comprises two key components: First, we construct a large…

Multimodal Reasoning

VideoRewardBench: Comprehensive Evaluation of Multimodal Reward Models for Video Understanding

2025-08-30 · Zhihong Zhang, Xiaojian Huang, Jin Xu, Zhuodong Luo 외 arxiv

Multimodal reward models (MRMs) play a crucial role in the training, inference, and evaluation of Large Vision Language Models (LVLMs) by assessing response quality. However, existing benchmarks for evaluating MRMs in th…

Reinforcement Learning

Med-RewardBench: Benchmarking Reward Models and Judges for Medical Multimodal Large Language Models

2025-08-29 · Meidan Ding, Jipeng Zhang, Wenxuan Wang, Cheng-Yi Li 외 arxiv

Multimodal large language models (MLLMs) hold significant potential in medical applications, including disease diagnosis and clinical decision-making. However, these tasks require highly accurate, context-sensitive, and …