English

LLaVA-Critic: Learning to Evaluate Multimodal Models

Computer Vision and Pattern Recognition 2025-03-05 v2 Computation and Language

Abstract

We introduce LLaVA-Critic, the first open-source large multimodal model (LMM) designed as a generalist evaluator to assess performance across a wide range of multimodal tasks. LLaVA-Critic is trained using a high-quality critic instruction-following dataset that incorporates diverse evaluation criteria and scenarios. Our experiments demonstrate the model's effectiveness in two key areas: (1) LMM-as-a-Judge, where LLaVA-Critic provides reliable evaluation scores, performing on par with or surpassing GPT models on multiple evaluation benchmarks; and (2) Preference Learning, where it generates reward signals for preference learning, enhancing model alignment capabilities. This work underscores the potential of open-source LMMs in self-critique and evaluation, setting the stage for future research into scalable, superhuman alignment feedback mechanisms for LMMs.

Keywords

Cite

@article{arxiv.2410.02712,
  title  = {LLaVA-Critic: Learning to Evaluate Multimodal Models},
  author = {Tianyi Xiong and Xiyao Wang and Dong Guo and Qinghao Ye and Haoqi Fan and Quanquan Gu and Heng Huang and Chunyuan Li},
  journal= {arXiv preprint arXiv:2410.02712},
  year   = {2025}
}

Comments

Accepted by CVPR 2025; Project Page: https://llava-vl.github.io/blog/2024-10-03-llava-critic

R2 v1 2026-06-28T19:07:23.433Z