CR³: Boosting Compositional Reasoning in MLLMs Through Rule-Based Reinforcement Learning

Shun Qian; Bingquan Liu; Chengjie Sun; Peijin Xie; Zhen Xu; Baoxun Wang

AAAI 2026technical0 citations

CR³: Boosting Compositional Reasoning in MLLMs Through Rule-Based Reinforcement Learning

Shun Qian, Bingquan Liu, Chengjie Sun, Peijin Xie, Zhen Xu, Baoxun Wang

Abstract

Compositional reasoning is a critical capability for multimodal models, enabling systematic understanding of complex scenes through structured combinations of objects, attributes, and relations. However, existing research on this ability primarily focuses on vision-language models (VLMs, e.g., CLIP and SigLIP), with limited exploration of multimodal large language models (MLLMs). To address this gap, we introduce CR³, a novel framework that enhances compositional reasoning abilities of MLLMs via rule-based reinforcement learning. CR³ leverages rule-based rewards to optimize the MLLM

BibTeX

@inproceedings{aaai2026_crboostingcompos,
  title = {CR³: Boosting Compositional Reasoning in MLLMs Through Rule-Based Reinforcement Learning},
  author = {Shun Qian and Bingquan Liu and Chengjie Sun and Peijin Xie and Zhen Xu and Baoxun Wang},
  booktitle = {AAAI 2026},
  year = {2026}
}