History-Guided Iterative Visual Reasoning with Self-Correction

TL;DR

The paper introduces H-GIVR, a novel iterative reasoning framework for multimodal large language models that dynamically corrects errors by reusing historical reasoning, significantly boosting accuracy across multiple datasets.

Contribution

H-GIVR enables models to actively correct visual understanding errors during iterative reasoning by reusing previous answers, unlike traditional fixed sampling methods.

Findings

Significant accuracy improvements on five datasets.

Low additional computational cost.

107% accuracy increase on ScienceQA with Llama3.2-vision:11b.

Abstract

Self-consistency methods are the core technique for improving the reasoning reliability of multimodal large language models (MLLMs). By generating multiple reasoning results through repeated sampling and selecting the best answer via voting, they play an important role in cross-modal tasks. However, most existing self-consistency methods are limited to a fixed ``repeated sampling and voting'' paradigm and do not reuse historical reasoning information. As a result, models struggle to actively correct visual understanding errors and dynamically adjust their reasoning during iteration. Inspired by the human reasoning behavior of repeated verification and dynamic error correction, we propose the H-GIVR framework. During iterative reasoning, the MLLM observes the image multiple times and uses previously generated answers as references for subsequent steps, enabling dynamic correction of errors and improving answer accuracy. We conduct comprehensive experiments on five datasets and three models. The results show that the H-GIVR framework can significantly improve cross-modal reasoning accuracy while maintaining low computational cost. For instance, using \texttt{Llama3.2-vision:11b} on the ScienceQA dataset, the model requires an average of 2.57 responses per question to achieve an accuracy of 78.90\%, representing a 107\% improvement over the baseline.