首个支持图文音频大模型的可解释性框架，让模型决策更透明。
mllm-shap: A Shapley Value Explainability Platform for Text-Audio Multimodal Large Language Models
- 针对多模态输入设计新型掩码策略，解决文本与音频混合处理难题。
- 通过语音对齐分组，使计算量降低10至50倍，支持长音频分析。
- 内置可视化界面，适合研究者和开发者调试多模态模型决策过程。
我们提出 mllm-shap，一个开源的 Python 框架，将基于 Shapley Value（SV）的可解释性从纯文本大语言模型拓展至处理文本与音频联合输入的多模态大语言模型（MLLM）。现有文本归因研究成熟，但 mllm-shap 解决了多模态场景下的三个关键挑战：(1) 模态感知的联盟掩码，有效管理离散文本词元与密集音频编码帧的交错处理；(2) 多轮对话追踪，利用每词元元数据维护角色与模态上下文；(3) 基于语音对齐的词元分组，将联盟空间缩小10至50倍，使长音频的 SV 估计在计算上可行。平台实现五种 SV 估计策略，包括具有奈曼最优分配的互补贡献（CC）估计器，在收敛速度上优于标准蒙特卡洛基线。mllm-shap 以 pip 可安装包形式提供，并配备交互式网页版可视化界面，用于细粒度归因分析。据我们所知，这是首个公开可用的、完整的、可复现的文本-音频多模态大模型基于 SV 的可解释性全流程框架。
原文摘要 · Abstract (English)
We introduce mllm-shap, an open-source Python framework designed to extend Shapley Value (SV) explainability from text-only Large Language Models to Multimodal LLMs (MLLMs) processing joint text and audio inputs. While text-based attribution is well-studied, mllm-shap addresses three critical challenges unique to the multimodal regime: (1) Modality-aware coalition masking, which manages the interleaved processing of discrete text tokens and dense audio encoder frames. (2) Multi-turn conversation tracking, utilizing per-token metadata to maintain role and modality context. (3) Phonetic alignment-based token grouping, a novel technique that reduces the coalition space by 10x to 50x, rendering SV estimation computationally feasible for long-form audio. The platform implements five SV estimation strategies, including a Complementary Contributions (CC) estimator with Neyman-optimal allocation that demonstrates superior convergence over standard Monte Carlo baselines. mllm-shap is provided as a pip-installable package featuring an interactive web-based GUI for granular attribution visualization. To our knowledge, this is the first publicly available framework providing a complete, reproducible pipeline for SV-based explainability in text-audio 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。