Context Attribution with Multi-Armed Bandit Optimization
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910155067621376 |
|---|---|
| author | Pan, Deng Murugesan, Keerthiram Hua, Ting Moniz, Nuno Chawla, Nitesh |
| author_facet | Pan, Deng Murugesan, Keerthiram Hua, Ting Moniz, Nuno Chawla, Nitesh |
| contents | Understanding which parts of the retrieved context contribute to a large language model's generated answer is essential for building interpretable and trustworthy retrieval-augmented generation. We propose a novel framework that formulates context attribution as a combinatorial multi-armed bandit problem. We utilize Linear Thompson Sampling to efficiently identify the most influential context segments while minimizing the number of model queries. Our reward function leverages token log-probabilities to measure how well a subset of segments supports the original response, making it applicable to both open-source and black-box API-based models. Unlike SHAP and other perturbation-based methods that sample subsets uniformly, our approach adaptively prioritizes informative subsets based on posterior estimates of segment relevance, reducing computational costs. Experiments on multiple QA benchmarks demonstrate that our method achieves up to 30\% reduction in model queries while matching or exceeding the attribution quality of existing approaches. Our code is publicly available at https://github.com/pd90506/camab. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_19977 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Context Attribution with Multi-Armed Bandit Optimization Pan, Deng Murugesan, Keerthiram Hua, Ting Moniz, Nuno Chawla, Nitesh Artificial Intelligence Understanding which parts of the retrieved context contribute to a large language model's generated answer is essential for building interpretable and trustworthy retrieval-augmented generation. We propose a novel framework that formulates context attribution as a combinatorial multi-armed bandit problem. We utilize Linear Thompson Sampling to efficiently identify the most influential context segments while minimizing the number of model queries. Our reward function leverages token log-probabilities to measure how well a subset of segments supports the original response, making it applicable to both open-source and black-box API-based models. Unlike SHAP and other perturbation-based methods that sample subsets uniformly, our approach adaptively prioritizes informative subsets based on posterior estimates of segment relevance, reducing computational costs. Experiments on multiple QA benchmarks demonstrate that our method achieves up to 30\% reduction in model queries while matching or exceeding the attribution quality of existing approaches. Our code is publicly available at https://github.com/pd90506/camab. |
| title | Context Attribution with Multi-Armed Bandit Optimization |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2506.19977 |