llmSHAP: A Principled Approach to LLM Explainability

📝 Original Info

• Title: llmSHAP: A Principled Approach to LLM Explainability
• ArXiv ID: 2511.01311
• Date: 2025-11-03
• Authors: ** 논문에 명시된 저자 정보가 제공되지 않았습니다. **

📝 Abstract

Feature attribution methods help make machine learning-based inference explainable by determining how much one or several features have contributed to a model's output. A particularly popular attribution method is based on the Shapley value from cooperative game theory, a measure that guarantees the satisfaction of several desirable principles, assuming deterministic inference. We apply the Shapley value to feature attribution in large language model (LLM)-based decision support systems, where inference is, by design, stochastic (non-deterministic). We then demonstrate when we can and cannot guarantee Shapley value principle satisfaction across different implementation variants applied to LLM-based decision support, and analyze how the stochastic nature of LLMs affects these guarantees. We also highlight trade-offs between explainable inference speed, agreement with exact Shapley value attributions, and principle attainment.

💡 Deep Analysis

📄 Full Content

Reference