AAAI 2026technical0 citations

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

Jinhao Chen, Zhen Yang, Jianxin Shi, Tianyu Wo, Jie Tang

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

BibTeX
@inproceedings{aaai2026_mathseimprovingm,
  title = {MathSE: Improving Multimodal Mathematical Reasoning via Self-Evolving Iterative Reflection and Reward-Guided Fine-Tuning},
  author = {Jinhao Chen and Zhen Yang and Jianxin Shi and Tianyu Wo and Jie Tang},
  booktitle = {AAAI 2026},
  year = {2026}
}
MathSE: Improving Multimodal Mathematical Reasoning via Self-Evolving Iterative Reflection and Reward-Guided Fine-Tuning · AAAI 2026