MIND: Multi-rationale INtegrated Discriminative Reasoning Framework for Multi-modal Large Models

Chuang Yu, Jinmiao Zhao, Mingxuan Zhao, Yunpeng Liu, Xiujun Shu, Yuanhao Feng, Bo Wang, Xiangyu Yue
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:151737-151756, 2026.

Abstract

Recently, multimodal large language models (MLLMs) have been widely applied to reasoning tasks. However, they suffer from limited multi-rationale semantic modeling, insufficient logical robustness, and susceptibility to misleading cues. Therefore, we propose a Multi-rationale INtegrated Discriminative (MIND) reasoning framework, which is designed to endow MLLMs with human-like cognitive abilities of “Understand $\rightarrow$ Rethink $\rightarrow$ Correct”, and achieves a paradigm evolution from passive imitation-based reasoning to active discriminative reasoning. Specifically, we introduce a Rationale Augmentation and Discrimination (RAD) paradigm, which provides a unified and extensible data foundation. Meanwhile, we design a Progressive Two-stage Correction Learning (P2CL) strategy. The first phase enhances multi-rationale positive learning, while the second phase enables active logic discrimination and correction. In addition, to mitigate representation entanglement in the multi-rationale semantic space, we propose a Multi-rationale Contrastive Alignment (MCA) optimization strategy. Extensive experiments show that our MIND achieves SOTA performance on multiple public datasets. Our data and code are available at https://github.com/YuChuang1205/MIND.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-yu26bv, title = {{MIND}: Multi-rationale {IN}tegrated Discriminative Reasoning Framework for Multi-modal Large Models}, author = {Yu, Chuang and Zhao, Jinmiao and Zhao, Mingxuan and Liu, Yunpeng and Shu, Xiujun and Feng, Yuanhao and Wang, Bo and Yue, Xiangyu}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {151737--151756}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/yu26bv/yu26bv.pdf}, url = {https://proceedings.mlr.press/v306/yu26bv.html}, abstract = {Recently, multimodal large language models (MLLMs) have been widely applied to reasoning tasks. However, they suffer from limited multi-rationale semantic modeling, insufficient logical robustness, and susceptibility to misleading cues. Therefore, we propose a Multi-rationale INtegrated Discriminative (MIND) reasoning framework, which is designed to endow MLLMs with human-like cognitive abilities of “Understand $\rightarrow$ Rethink $\rightarrow$ Correct”, and achieves a paradigm evolution from passive imitation-based reasoning to active discriminative reasoning. Specifically, we introduce a Rationale Augmentation and Discrimination (RAD) paradigm, which provides a unified and extensible data foundation. Meanwhile, we design a Progressive Two-stage Correction Learning (P2CL) strategy. The first phase enhances multi-rationale positive learning, while the second phase enables active logic discrimination and correction. In addition, to mitigate representation entanglement in the multi-rationale semantic space, we propose a Multi-rationale Contrastive Alignment (MCA) optimization strategy. Extensive experiments show that our MIND achieves SOTA performance on multiple public datasets. Our data and code are available at https://github.com/YuChuang1205/MIND.} }
Endnote
%0 Conference Paper %T MIND: Multi-rationale INtegrated Discriminative Reasoning Framework for Multi-modal Large Models %A Chuang Yu %A Jinmiao Zhao %A Mingxuan Zhao %A Yunpeng Liu %A Xiujun Shu %A Yuanhao Feng %A Bo Wang %A Xiangyu Yue %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-yu26bv %I PMLR %P 151737--151756 %U https://proceedings.mlr.press/v306/yu26bv.html %V 306 %X Recently, multimodal large language models (MLLMs) have been widely applied to reasoning tasks. However, they suffer from limited multi-rationale semantic modeling, insufficient logical robustness, and susceptibility to misleading cues. Therefore, we propose a Multi-rationale INtegrated Discriminative (MIND) reasoning framework, which is designed to endow MLLMs with human-like cognitive abilities of “Understand $\rightarrow$ Rethink $\rightarrow$ Correct”, and achieves a paradigm evolution from passive imitation-based reasoning to active discriminative reasoning. Specifically, we introduce a Rationale Augmentation and Discrimination (RAD) paradigm, which provides a unified and extensible data foundation. Meanwhile, we design a Progressive Two-stage Correction Learning (P2CL) strategy. The first phase enhances multi-rationale positive learning, while the second phase enables active logic discrimination and correction. In addition, to mitigate representation entanglement in the multi-rationale semantic space, we propose a Multi-rationale Contrastive Alignment (MCA) optimization strategy. Extensive experiments show that our MIND achieves SOTA performance on multiple public datasets. Our data and code are available at https://github.com/YuChuang1205/MIND.
APA
Yu, C., Zhao, J., Zhao, M., Liu, Y., Shu, X., Feng, Y., Wang, B. & Yue, X.. (2026). MIND: Multi-rationale INtegrated Discriminative Reasoning Framework for Multi-modal Large Models. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:151737-151756 Available from https://proceedings.mlr.press/v306/yu26bv.html.

Related Material