Guardado en:
| Autores principales: | , , , , , , , , , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | https://arxiv.org/abs/2503.06252 |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866911087333474304 |
|---|---|
| author | Xiang, Kun Liu, Zhili Jiang, Zihao Nie, Yunshuang Cai, Kaixin Yin, Yiyang Huang, Runhui Fan, Haoxiang Li, Hanhui Huang, Weiran Zeng, Yihan Yuan, Yu-Jie Han, Jianhua Hong, Lanqing Xu, Hang Liang, Xiaodan |
| author_facet | Xiang, Kun Liu, Zhili Jiang, Zihao Nie, Yunshuang Cai, Kaixin Yin, Yiyang Huang, Runhui Fan, Haoxiang Li, Hanhui Huang, Weiran Zeng, Yihan Yuan, Yu-Jie Han, Jianhua Hong, Lanqing Xu, Hang Liang, Xiaodan |
| contents | In this paper, we address the challenging task of multimodal mathematical reasoning by incorporating the ability of "slow thinking" into multimodal large language models (MLLMs). Our core idea is that different levels of reasoning abilities can be combined dynamically to tackle questions with different complexity. To this end, we propose a paradigm of Self-structured Chain of Thought (SCoT), which is composed of minimal semantic atomic steps. Different from existing methods that rely on structured templates or free-form paradigms, our method can not only generate cognitive CoT structures for various complex tasks but also mitigates the phenomenon of overthinking. To introduce structured reasoning capabilities into visual understanding models, we further design a novel AtomThink framework with four key modules, including (i) a data engine to generate high-quality multimodal reasoning paths; (ii) a supervised fine-tuning process with serialized inference data; (iii) a policy-guided multi-turn inference method; and (iv) an atomic capability metric to evaluate the single step utilization rate. We conduct extensive experiments to show that the proposed AtomThink significantly improves the performance of baseline MLLMs, achieving more than 10\% average accuracy gains on MathVista and MathVerse. Compared to state-of-the-art structured CoT approaches, our method not only achieves higher accuracy but also improves data utilization by 5 times and boosts inference efficiency by 85.3\%. Our code is now public available in https://github.com/Quinn777/AtomThink. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_06252 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Can Atomic Step Decomposition Enhance the Self-structured Reasoning of Multimodal Large Models? Xiang, Kun Liu, Zhili Jiang, Zihao Nie, Yunshuang Cai, Kaixin Yin, Yiyang Huang, Runhui Fan, Haoxiang Li, Hanhui Huang, Weiran Zeng, Yihan Yuan, Yu-Jie Han, Jianhua Hong, Lanqing Xu, Hang Liang, Xiaodan Computer Vision and Pattern Recognition Artificial Intelligence In this paper, we address the challenging task of multimodal mathematical reasoning by incorporating the ability of "slow thinking" into multimodal large language models (MLLMs). Our core idea is that different levels of reasoning abilities can be combined dynamically to tackle questions with different complexity. To this end, we propose a paradigm of Self-structured Chain of Thought (SCoT), which is composed of minimal semantic atomic steps. Different from existing methods that rely on structured templates or free-form paradigms, our method can not only generate cognitive CoT structures for various complex tasks but also mitigates the phenomenon of overthinking. To introduce structured reasoning capabilities into visual understanding models, we further design a novel AtomThink framework with four key modules, including (i) a data engine to generate high-quality multimodal reasoning paths; (ii) a supervised fine-tuning process with serialized inference data; (iii) a policy-guided multi-turn inference method; and (iv) an atomic capability metric to evaluate the single step utilization rate. We conduct extensive experiments to show that the proposed AtomThink significantly improves the performance of baseline MLLMs, achieving more than 10\% average accuracy gains on MathVista and MathVerse. Compared to state-of-the-art structured CoT approaches, our method not only achieves higher accuracy but also improves data utilization by 5 times and boosts inference efficiency by 85.3\%. Our code is now public available in https://github.com/Quinn777/AtomThink. |
| title | Can Atomic Step Decomposition Enhance the Self-structured Reasoning of Multimodal Large Models? |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2503.06252 |