Score: 1

TokenShapley: Token Level Context Attribution with Shapley Value

Published: June 18, 2025 | arXiv ID: 2507.05261v2

By: Yingtai Xiao , Yuqing Zhu , Sirat Samyoun and more

Potential Business Impact:

Shows where AI got its facts from.

Business Areas:
Semantic Search Internet Services

Large language models (LLMs) demonstrate strong capabilities in in-context learning, but verifying the correctness of their generated responses remains a challenge. Prior work has explored attribution at the sentence level, but these methods fall short when users seek attribution for specific keywords within the response, such as numbers, years, or names. To address this limitation, we propose TokenShapley, a novel token-level attribution method that combines Shapley value-based data attribution with KNN-based retrieval techniques inspired by recent advances in KNN-augmented LLMs. By leveraging a precomputed datastore for contextual retrieval and computing Shapley values to quantify token importance, TokenShapley provides a fine-grained data attribution approach. Extensive evaluations on four benchmarks show that TokenShapley outperforms state-of-the-art baselines in token-level attribution, achieving an 11-23% improvement in accuracy.

Page Count
13 pages

Category
Computer Science:
Computation and Language