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

TL;DR

The paper introduces MIND, a novel reasoning framework for multimodal large language models that enhances multi-rationale semantic understanding, logical robustness, and active correction capabilities, leading to state-of-the-art results.

Contribution

It presents a comprehensive framework combining rationale augmentation, progressive correction, and contrastive alignment to improve reasoning in multimodal large models.

Findings

Achieves SOTA performance on multiple reasoning datasets.

Enhances logical robustness and multi-rationale semantic modeling.

Provides a new paradigm for active discriminative reasoning in MLLMs.

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 are susceptible to misleading interpretations in complex scenarios. Therefore, we propose a Multi-rationale INtegrated Discriminative (MIND) reasoning framework, which is designed to endow MLLMs with human-like cognitive abilities of "Understand -> Rethink -> 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 automatically and efficiently expands existing datasets by generating diverse rationales, providing 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, which achieves semantic aggregation of correct reasoning and boundary separation of incorrect reasoning. Extensive experiments demonstrate that the proposed MIND reasoning framework achieves state-of-the-art (SOTA) performance on multiple public datasets covering scientific, commonsense, and mathematical scenarios. It provides a new perspective for advancing MLLMs towards higher levels of cognitive intelligence. Our code is available at https://github.com/YuChuang1205/MIND