{"ID":2846551,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2511.01311","arxiv_id":"2511.01311","title":"llmSHAP: A Principled Approach to LLM Explainability","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.","short_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 sev...","url_abs":"https://arxiv.org/abs/2511.01311","url_pdf":"https://arxiv.org/pdf/2511.01311v1","authors":"[\"Filip Naudot\",\"Tobias Sundqvist\",\"Timotheus Kampik\"]","published":"2025-11-03T07:54:47Z","proceeding":"cs.AI","tasks":"[\"cs.AI\"]","methods":"[\"Large Language Model\",\"Language Model\"]","has_code":false}