MME-CoT: Benchmarking Chain-of-Thought in Large Multimodal Models for Reasoning Quality, Robustness, and Efficiency

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Jiang, Dongzhi, Zhang, Renrui, Guo, Ziyu, Li, Yanwei, Qi, Yu, Chen, Xinyan, Wang, Liuhui, Jin, Jianhan, Guo, Claire, Yan, Shen, Zhang, Bo, Fu, Chaoyou, Gao, Peng, Li, Hongsheng
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910825907748864
author Jiang, Dongzhi
Zhang, Renrui
Guo, Ziyu
Li, Yanwei
Qi, Yu
Chen, Xinyan
Wang, Liuhui
Jin, Jianhan
Guo, Claire
Yan, Shen
Zhang, Bo
Fu, Chaoyou
Gao, Peng
Li, Hongsheng
author_facet Jiang, Dongzhi
Zhang, Renrui
Guo, Ziyu
Li, Yanwei
Qi, Yu
Chen, Xinyan
Wang, Liuhui
Jin, Jianhan
Guo, Claire
Yan, Shen
Zhang, Bo
Fu, Chaoyou
Gao, Peng
Li, Hongsheng
contents Answering questions with Chain-of-Thought (CoT) has significantly enhanced the reasoning capabilities of Large Language Models (LLMs), yet its impact on Large Multimodal Models (LMMs) still lacks a systematic assessment and in-depth investigation. In this paper, we introduce MME-CoT, a specialized benchmark evaluating the CoT reasoning performance of LMMs, spanning six domains: math, science, OCR, logic, space-time, and general scenes. As the first comprehensive study in this area, we propose a thorough evaluation suite incorporating three novel metrics that assess the reasoning quality, robustness, and efficiency at a fine-grained level. Leveraging curated high-quality data and a unique evaluation strategy, we conduct an in-depth analysis of state-of-the-art LMMs, uncovering several key insights: 1) Models with reflection mechanism demonstrate a superior CoT quality, with Kimi k1.5 outperforming GPT-4o and demonstrating the highest quality results; 2) CoT prompting often degrades LMM performance on perception-heavy tasks, suggesting a potentially harmful overthinking behavior; and 3) Although the CoT quality is high, LMMs with reflection exhibit significant inefficiency in both normal response and self-correction phases. We hope MME-CoT serves as a foundation for advancing multimodal reasoning in LMMs. Project Page: https://mmecot.github.io/
format Preprint
id arxiv_https___arxiv_org_abs_2502_09621
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle MME-CoT: Benchmarking Chain-of-Thought in Large Multimodal Models for Reasoning Quality, Robustness, and Efficiency
Jiang, Dongzhi
Zhang, Renrui
Guo, Ziyu
Li, Yanwei
Qi, Yu
Chen, Xinyan
Wang, Liuhui
Jin, Jianhan
Guo, Claire
Yan, Shen
Zhang, Bo
Fu, Chaoyou
Gao, Peng
Li, Hongsheng
Computer Vision and Pattern Recognition
Artificial Intelligence
Computation and Language
Answering questions with Chain-of-Thought (CoT) has significantly enhanced the reasoning capabilities of Large Language Models (LLMs), yet its impact on Large Multimodal Models (LMMs) still lacks a systematic assessment and in-depth investigation. In this paper, we introduce MME-CoT, a specialized benchmark evaluating the CoT reasoning performance of LMMs, spanning six domains: math, science, OCR, logic, space-time, and general scenes. As the first comprehensive study in this area, we propose a thorough evaluation suite incorporating three novel metrics that assess the reasoning quality, robustness, and efficiency at a fine-grained level. Leveraging curated high-quality data and a unique evaluation strategy, we conduct an in-depth analysis of state-of-the-art LMMs, uncovering several key insights: 1) Models with reflection mechanism demonstrate a superior CoT quality, with Kimi k1.5 outperforming GPT-4o and demonstrating the highest quality results; 2) CoT prompting often degrades LMM performance on perception-heavy tasks, suggesting a potentially harmful overthinking behavior; and 3) Although the CoT quality is high, LMMs with reflection exhibit significant inefficiency in both normal response and self-correction phases. We hope MME-CoT serves as a foundation for advancing multimodal reasoning in LMMs. Project Page: https://mmecot.github.io/
title MME-CoT: Benchmarking Chain-of-Thought in Large Multimodal Models for Reasoning Quality, Robustness, and Efficiency
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2502.09621