RAGferee: Building Contextual Reward Models for Retrieval-Augmented Generation
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_ | 1866908568671748096 |
|---|---|
| author | Coman, Andrei C. Sorodoc, Ionut-Teodor Ribeiro, Leonardo F. R. Byrne, Bill Henderson, James de Gispert, Adrià |
| author_facet | Coman, Andrei C. Sorodoc, Ionut-Teodor Ribeiro, Leonardo F. R. Byrne, Bill Henderson, James de Gispert, Adrià |
| contents | Existing Reward Models (RMs), typically trained on general preference data, struggle in Retrieval Augmented Generation (RAG) settings, which require judging responses for faithfulness to retrieved context, relevance to the user query, appropriate refusals when context is insufficient, completeness and conciseness of information. To address the lack of publicly available RAG-centric preference datasets and specialised RMs, we introduce RAGferee, a methodology that repurposes question-answering (QA) datasets into preference pairs that prioritise groundedness over stylistic features, enabling the training of contextual RMs better suited to judging RAG responses. Using RAGferee, we curate a small preference dataset of 4K samples and fine-tune RMs ranging from 7B to 24B parameters. Our RAG-centric RMs achieve state-of-the-art performance on ContextualJudgeBench, surpassing existing 70B+ RMs trained on much larger (up to 2.4M samples) general corpora, with an absolute improvement of +15.5%. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_26011 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | RAGferee: Building Contextual Reward Models for Retrieval-Augmented Generation Coman, Andrei C. Sorodoc, Ionut-Teodor Ribeiro, Leonardo F. R. Byrne, Bill Henderson, James de Gispert, Adrià Computation and Language Existing Reward Models (RMs), typically trained on general preference data, struggle in Retrieval Augmented Generation (RAG) settings, which require judging responses for faithfulness to retrieved context, relevance to the user query, appropriate refusals when context is insufficient, completeness and conciseness of information. To address the lack of publicly available RAG-centric preference datasets and specialised RMs, we introduce RAGferee, a methodology that repurposes question-answering (QA) datasets into preference pairs that prioritise groundedness over stylistic features, enabling the training of contextual RMs better suited to judging RAG responses. Using RAGferee, we curate a small preference dataset of 4K samples and fine-tune RMs ranging from 7B to 24B parameters. Our RAG-centric RMs achieve state-of-the-art performance on ContextualJudgeBench, surpassing existing 70B+ RMs trained on much larger (up to 2.4M samples) general corpora, with an absolute improvement of +15.5%. |
| title | RAGferee: Building Contextual Reward Models for Retrieval-Augmented Generation |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2509.26011 |