Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2509.15839 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908547203203072 |
|---|---|
| author | Luo, Zhongze Yin, Zhenshuai Guo, Yongxin Wang, Zhichao Zhu, Jionghao Tang, Xiaoying |
| author_facet | Luo, Zhongze Yin, Zhenshuai Guo, Yongxin Wang, Zhichao Zhu, Jionghao Tang, Xiaoying |
| contents | While multimodal LLMs (MLLMs) demonstrate remarkable reasoning progress, their application in specialized scientific domains like physics reveals significant gaps in current evaluation benchmarks. Specifically, existing benchmarks often lack fine-grained subject coverage, neglect the step-by-step reasoning process, and are predominantly English-centric, failing to systematically evaluate the role of visual information. Therefore, we introduce \textbf {Multi-Physics} for Chinese physics reasoning, a comprehensive benchmark that includes 5 difficulty levels, featuring 1,412 image-associated, multiple-choice questions spanning 11 high-school physics subjects. We employ a dual evaluation framework to evaluate 20 different MLLMs, analyzing both final answer accuracy and the step-by-step integrity of their chain-of-thought. Furthermore, we systematically study the impact of difficulty level and visual information by comparing the model performance before and after changing the input mode. Our work provides not only a fine-grained resource for the community but also offers a robust methodology for dissecting the multimodal reasoning process of state-of-the-art MLLMs, and our dataset and code have been open-sourced: https://github.com/luozhongze/Multi-Physics. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_15839 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Multi-Physics: A Comprehensive Benchmark for Multimodal LLMs Reasoning on Chinese Multi-Subject Physics Problems Luo, Zhongze Yin, Zhenshuai Guo, Yongxin Wang, Zhichao Zhu, Jionghao Tang, Xiaoying Computation and Language While multimodal LLMs (MLLMs) demonstrate remarkable reasoning progress, their application in specialized scientific domains like physics reveals significant gaps in current evaluation benchmarks. Specifically, existing benchmarks often lack fine-grained subject coverage, neglect the step-by-step reasoning process, and are predominantly English-centric, failing to systematically evaluate the role of visual information. Therefore, we introduce \textbf {Multi-Physics} for Chinese physics reasoning, a comprehensive benchmark that includes 5 difficulty levels, featuring 1,412 image-associated, multiple-choice questions spanning 11 high-school physics subjects. We employ a dual evaluation framework to evaluate 20 different MLLMs, analyzing both final answer accuracy and the step-by-step integrity of their chain-of-thought. Furthermore, we systematically study the impact of difficulty level and visual information by comparing the model performance before and after changing the input mode. Our work provides not only a fine-grained resource for the community but also offers a robust methodology for dissecting the multimodal reasoning process of state-of-the-art MLLMs, and our dataset and code have been open-sourced: https://github.com/luozhongze/Multi-Physics. |
| title | Multi-Physics: A Comprehensive Benchmark for Multimodal LLMs Reasoning on Chinese Multi-Subject Physics Problems |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2509.15839 |