揭示多模态大模型生成文本时依赖的视觉区域与语言先验
Where MLLMs Attend and What They Rely On: Explaining Autoregressive Token Generation
- 通过稀疏图像区域贪婪搜索，量化视觉与语言对生成的影响
- 在多个开源模型上验证，显著提升定位精度与幻觉诊断能力
- 轻量级设计，显存占用少，适合实际部署解释需求
多模态大语言模型（MLLM）在将视觉输入与自然语言输出对齐方面表现出色，但其生成的词语在多大程度上依赖视觉信息仍不清楚，限制了模型的可解释性与可靠性。本文提出EAGLE，一种轻量级黑箱框架，用于解释MLLM中自回归标记生成过程。EAGLE将选定标记归因于紧凑的感知区域，并量化语言先验与感知证据的相对影响。该框架引入统一的充分性（洞察分数）与必要性（必要分数）目标函数，通过稀疏化图像区域的贪婪搜索进行优化，实现忠实且高效的归因。除空间归因外，EAGLE还进行模态感知分析，分离不同标记所依赖的模态，提供细粒度的模型决策解释。在多个开源MLLM上的大量实验表明，EAGLE在忠实性、定位准确性和幻觉诊断方面均优于现有方法，同时显著降低GPU内存消耗。结果证明其在提升MLLM可解释性方面的有效性与实用性。
原文摘要 · Abstract (English)
Multimodal large language models (MLLMs) have demonstrated remarkable capabilities in aligning visual inputs with natural language outputs. Yet, the extent to which generated tokens depend on visual modalities remains poorly understood, limiting interpretability and reliability. In this work, we present EAGLE, a lightweight black-box framework for explaining autoregressive token generation in MLLMs. EAGLE attributes any selected tokens to compact perceptual regions while quantifying the relative influence of language priors and perceptual evidence. The framework introduces an objective function that unifies sufficiency (insight score) and indispensability (necessity score), optimized via greedy search over sparsified image regions for faithful and efficient attribution. Beyond spatial attribution, EAGLE performs modality-aware analysis that disentangles what tokens rely on, providing fine-grained interpretability of model decisions. Extensive experiments across open-source MLLMs show that EAGLE consistently outperforms existing methods in faithfulness, localization, and hallucination diagnosis, while requiring substantially less GPU memory. These results highlight its effectiveness and practicality for advancing the interpretability of MLLMs.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。