LLaVA-Critic: Learning to Evaluate Multimodal Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Xiong, Tianyi, Wang, Xiyao, Guo, Dong, Ye, Qinghao, Fan, Haoqi, Gu, Quanquan, Huang, Heng, Li, Chunyuan
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913717346631680
author Xiong, Tianyi
Wang, Xiyao
Guo, Dong
Ye, Qinghao
Fan, Haoqi
Gu, Quanquan
Huang, Heng
Li, Chunyuan
author_facet Xiong, Tianyi
Wang, Xiyao
Guo, Dong
Ye, Qinghao
Fan, Haoqi
Gu, Quanquan
Huang, Heng
Li, Chunyuan
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2410_02712
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle LLaVA-Critic: Learning to Evaluate Multimodal Models
Xiong, Tianyi
Wang, Xiyao
Guo, Dong
Ye, Qinghao
Fan, Haoqi
Gu, Quanquan
Huang, Heng
Li, Chunyuan
Computer Vision and Pattern Recognition
Computation and Language
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.
title LLaVA-Critic: Learning to Evaluate Multimodal Models
topic Computer Vision and Pattern Recognition
Computation and Language
url https://arxiv.org/abs/2410.02712