Mitigating Visual Forgetting via Take-along Visual Conditioning for Multi-modal Long CoT Reasoning
This addresses a specific bottleneck in multimodal reasoning for tasks like geometry problems, offering an incremental improvement over existing methods.
The paper tackles the problem of visual forgetting in multimodal LLMs during long reasoning chains, where models lose focus on visual information over time, and proposes Take-along Visual Conditioning (TVC) to shift and compress visual inputs, achieving state-of-the-art performance with a +3.4 point average gain on five mathematical reasoning benchmarks.
Recent advancements in Large Language Models (LLMs) have demonstrated enhanced reasoning capabilities, evolving from Chain-of-Thought (CoT) prompting to advanced, product-oriented solutions like OpenAI o1. During our re-implementation of this model, we noticed that in multimodal tasks requiring visual input (e.g., geometry problems), Multimodal LLMs (MLLMs) struggle to maintain focus on the visual information, in other words, MLLMs suffer from a gradual decline in attention to visual information as reasoning progresses, causing text-over-relied outputs. To investigate this, we ablate image inputs during long-chain reasoning. Concretely, we truncate the reasoning process midway, then re-complete the reasoning process with the input image removed. We observe only a ~2% accuracy drop on MathVista's test-hard subset, revealing the model's textual outputs dominate the following reasoning process. Motivated by this, we propose Take-along Visual Conditioning (TVC), a strategy that shifts image input to critical reasoning stages and compresses redundant visual tokens via dynamic pruning. This methodology helps the model retain attention to the visual components throughout the reasoning. Our approach achieves state-of-the-art performance on average across five mathematical reasoning benchmarks (+3.4 points vs previous sota), demonstrating the effectiveness of TVC in enhancing multimodal reasoning systems.