Adapting Large Language Models for Multi-Domain Retrieval-Augmented-Generation

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Misrahi, Alexandre, Chirkova, Nadezhda, Louis, Maxime, Nikoulina, Vassilina
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866915225469452288
author Misrahi, Alexandre
Chirkova, Nadezhda
Louis, Maxime
Nikoulina, Vassilina
author_facet Misrahi, Alexandre
Chirkova, Nadezhda
Louis, Maxime
Nikoulina, Vassilina
contents Retrieval-Augmented Generation (RAG) enhances LLM factuality, but multi-domain applications face challenges like lack of diverse benchmarks and poor out-of-domain generalization. The first contribution of this work is to introduce a diverse benchmark comprising a variety of question-answering tasks from 8 sources and covering 13 domains. Our second contribution consists in systematically testing out-of-domain generalization for typical RAG tuning strategies. While our findings reveal that standard fine-tuning fails to generalize effectively, we show that sequence-level distillation with teacher-generated labels improves out-of-domain performance by providing more coherent supervision. Our findings highlight key strategies for improving multi-domain RAG robustness.
format Preprint
id arxiv_https___arxiv_org_abs_2504_02411
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Adapting Large Language Models for Multi-Domain Retrieval-Augmented-Generation
Misrahi, Alexandre
Chirkova, Nadezhda
Louis, Maxime
Nikoulina, Vassilina
Computation and Language
Retrieval-Augmented Generation (RAG) enhances LLM factuality, but multi-domain applications face challenges like lack of diverse benchmarks and poor out-of-domain generalization. The first contribution of this work is to introduce a diverse benchmark comprising a variety of question-answering tasks from 8 sources and covering 13 domains. Our second contribution consists in systematically testing out-of-domain generalization for typical RAG tuning strategies. While our findings reveal that standard fine-tuning fails to generalize effectively, we show that sequence-level distillation with teacher-generated labels improves out-of-domain performance by providing more coherent supervision. Our findings highlight key strategies for improving multi-domain RAG robustness.
title Adapting Large Language Models for Multi-Domain Retrieval-Augmented-Generation
topic Computation and Language
url https://arxiv.org/abs/2504.02411