[2603.26348] Reflect to Inform: Boosting Multimodal Reasoning via Information-Gain-Driven Verification

Computer Science > Computer Vision and Pattern Recognition arXiv:2603.26348 (cs) [Submitted on 27 Mar 2026] Title:Reflect to Inform: Boosting Multimodal Reasoning via Information-Gain-Driven Verification Authors:Shuai Lv, Chang Liu, Feng Tang, Yujie Yuan, Aojun Zhou, Kui Zhang, Xi Yang, Yangqiu Song View a PDF of the paper titled Reflect to Inform: Boosting Multimodal Reasoning via Information-Gain-Driven Verification, by Shuai Lv and 7 other authors View PDF HTML (experimental) Abstract:Multimodal Large Language Models (MLLMs) achieve strong multimodal reasoning performance, yet we identify a recurring failure mode in long-form generation: as outputs grow longer, models progressively drift away from image evidence and fall back on textual priors, resulting in ungrounded reasoning and hallucinations. Interestingly, Based on attention analysis, we find that MLLMs have a latent capability for late-stage visual verification that is present but not consistently activated. Motivated by this observation, we propose Visual Re-Examination (VRE), a self-evolving training framework that enables MLLMs to autonomously perform visual introspection during reasoning without additional visual inputs. Rather than distilling visual capabilities from a stronger teacher, VRE promotes iterative self-improvement by leveraging the model itself to generate reflection traces, making visual information actionable through information gain. Extensive experiments across diverse multimodal benchmarks d...