用可解释的加性模型即时计算沙普利值，提升效率与可读性。
InstaSHAP: Interpretable Additive Models Explain Shapley Values Instantly
- 从变分视角建立加性模型与沙普利值的联系，指导新方法设计。
- 训练出的模型可单次前向传播直接输出沙普利值，速度显著提升。
- 揭示加性模型表达能力局限，为沙普利值在视觉/自然语言任务中的应用提供反思。
近年来，沙普利值与SHAP解释已成为黑箱模型事后解释的主要范式。尽管其理论基础坚实，但诸多研究关注其计算效率与表达能力的局限。当前文献中，其与加性模型的深层关联却被严重忽视。本文发现，从变分视角出发，将广义加性模型（GAM）与SHAP解释相联系，能深入揭示近来多数进展的本质。基于此，我们反向借用该思想，提出一种新方法：训练可解释的GAM模型，使其自动优化以实现沙普利值的单次前向传播计算。最后，我们给出理论结果，表明GAM模型的表达能力受限，正是当前SHAP方法在计算机视觉与自然语言处理中应用时的核心短板。
原文摘要 · Abstract (English)
In recent years, the Shapley value and SHAP explanations have emerged as one of the most dominant paradigms for providing post-hoc explanations of black-box models. Despite their well-founded theoretical properties, many recent works have focused on the limitations in both their computational efficiency and their representation power. The underlying connection with additive models, however, is left critically under-emphasized in the current literature. In this work, we find that a variational perspective linking GAM models and SHAP explanations is able to provide deep insights into nearly all recent developments. In light of this connection, we borrow in the other direction to develop a new method to train interpretable GAM models which are automatically purified to compute the Shapley value in a single forward pass. Finally, we provide theoretical results showing the limited representation power of GAM models is the same Achilles' heel existing in SHAP and discuss the implications for SHAP's modern usage in CV and NLP.
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。