Towards End-to-End Model-Agnostic Explanations for RAG Systems
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_ | 1866914029854785536 |
|---|---|
| author | Sudhi, Viju Bhat, Sinchana Ramakanth Rudat, Max Teucher, Roman Flores-Herr, Nicolas |
| author_facet | Sudhi, Viju Bhat, Sinchana Ramakanth Rudat, Max Teucher, Roman Flores-Herr, Nicolas |
| contents | Retrieval Augmented Generation (RAG) systems, despite their growing popularity for enhancing model response reliability, often struggle with trustworthiness and explainability. In this work, we present a novel, holistic, model-agnostic, post-hoc explanation framework leveraging perturbation-based techniques to explain the retrieval and generation processes in a RAG system. We propose different strategies to evaluate these explanations and discuss the sufficiency of model-agnostic explanations in RAG systems. With this work, we further aim to catalyze a collaborative effort to build reliable and explainable RAG systems. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_07620 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Towards End-to-End Model-Agnostic Explanations for RAG Systems Sudhi, Viju Bhat, Sinchana Ramakanth Rudat, Max Teucher, Roman Flores-Herr, Nicolas Information Retrieval Retrieval Augmented Generation (RAG) systems, despite their growing popularity for enhancing model response reliability, often struggle with trustworthiness and explainability. In this work, we present a novel, holistic, model-agnostic, post-hoc explanation framework leveraging perturbation-based techniques to explain the retrieval and generation processes in a RAG system. We propose different strategies to evaluate these explanations and discuss the sufficiency of model-agnostic explanations in RAG systems. With this work, we further aim to catalyze a collaborative effort to build reliable and explainable RAG systems. |
| title | Towards End-to-End Model-Agnostic Explanations for RAG Systems |
| topic | Information Retrieval |
| url | https://arxiv.org/abs/2509.07620 |