MathSE: Improving Multimodal Mathematical Reasoning via Self-Evolving Iterative Reflection and Reward-Guided Fine-Tuning

TL;DR

MathSE introduces a self-evolving iterative reflection and reward-guided fine-tuning framework that enhances multimodal mathematical reasoning by iteratively refining the model through inference, reflection, and feedback, surpassing existing models.

Contribution

It proposes a novel iterative fine-tuning framework for multimodal models that improves reasoning by incorporating reflection and reward feedback, addressing limitations of static datasets.

Findings

Significant performance improvements on mathematical reasoning benchmarks.

Outperforms leading open-source multimodal mathematical reasoning models.

Effective in handling complex and novel mathematical questions.

Abstract

Multimodal large language models (MLLMs) have demonstrated remarkable capabilities in vision-language answering tasks. Despite their strengths, these models often encounter challenges in achieving complex reasoning tasks such as mathematical problem-solving. Previous works have focused on fine-tuning on specialized mathematical datasets. However, these datasets are typically distilled directly from teacher models, which capture only static reasoning patterns and leaving substantial gaps compared to student models. This reliance on fixed teacher-derived datasets not only restricts the model's ability to adapt to novel or more intricate questions that extend beyond the confines of the training data, but also lacks the iterative depth needed for robust generalization. To overcome these limitations, we propose \textbf{\method}, a \textbf{Math}ematical \textbf{S}elf-\textbf{E}volving framework for MLLMs. In contrast to traditional one-shot fine-tuning paradigms, \method iteratively refines the model through cycles of inference, reflection, and reward-based feedback. Specifically, we leverage iterative fine-tuning by incorporating correct reasoning paths derived from previous-stage inference and integrating reflections from a specialized Outcome Reward Model (ORM). To verify the effectiveness of \method, we evaluate it on a suite of challenging benchmarks, demonstrating significant performance gains over backbone models. Notably, our experimental results on MathVL-test surpass the leading open-source multimodal mathematical reasoning model QVQ. Our code and models are available at \texttt{https://zheny2751\allowbreak-dotcom.github.io/\allowbreak MathSE.github.io/}.