DRAG: Distilling RAG for SLMs from LLMs to Transfer Knowledge and Mitigate Hallucination via Evidence and Graph-based Distillation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chen, Jennifer, Myrzakhan, Aidar, Luo, Yaxin, Khan, Hassaan Muhammad, Bsharat, Sondos Mahmoud, Shen, Zhiqiang
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908389796216832
author Chen, Jennifer
Myrzakhan, Aidar
Luo, Yaxin
Khan, Hassaan Muhammad
Bsharat, Sondos Mahmoud
Shen, Zhiqiang
author_facet Chen, Jennifer
Myrzakhan, Aidar
Luo, Yaxin
Khan, Hassaan Muhammad
Bsharat, Sondos Mahmoud
Shen, Zhiqiang
contents Retrieval-Augmented Generation (RAG) methods have proven highly effective for tasks requiring factual consistency and robust knowledge retrieval. However, large-scale RAG systems consume significant computational resources and are prone to generating hallucinated content from Humans. In this work, we introduce $\texttt{DRAG}$, a novel framework for distilling RAG knowledge from large-scale Language Models (LLMs) into small LMs (SLMs). Our approach leverages evidence- and knowledge graph-based distillation, ensuring that the distilled model retains critical factual knowledge while significantly reducing model size and computational cost. By aligning the smaller model's predictions with a structured knowledge graph and ranked evidence, $\texttt{DRAG}$ effectively mitigates hallucinations and improves factual accuracy. We further present a case demonstrating how our framework mitigates user privacy risks and introduce a corresponding benchmark. Experimental evaluations on multiple benchmarks demonstrate that our method outperforms the prior competitive RAG methods like MiniRAG for SLMs by up to 27.7% using the same models, preserving high-level efficiency and reliability. With $\texttt{DRAG}$, we provide a practical and resource-efficient roadmap to deploying enhanced retrieval and generation capabilities in small-sized LLMs.
format Preprint
id arxiv_https___arxiv_org_abs_2506_01954
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle DRAG: Distilling RAG for SLMs from LLMs to Transfer Knowledge and Mitigate Hallucination via Evidence and Graph-based Distillation
Chen, Jennifer
Myrzakhan, Aidar
Luo, Yaxin
Khan, Hassaan Muhammad
Bsharat, Sondos Mahmoud
Shen, Zhiqiang
Computation and Language
Artificial Intelligence
Machine Learning
Retrieval-Augmented Generation (RAG) methods have proven highly effective for tasks requiring factual consistency and robust knowledge retrieval. However, large-scale RAG systems consume significant computational resources and are prone to generating hallucinated content from Humans. In this work, we introduce $\texttt{DRAG}$, a novel framework for distilling RAG knowledge from large-scale Language Models (LLMs) into small LMs (SLMs). Our approach leverages evidence- and knowledge graph-based distillation, ensuring that the distilled model retains critical factual knowledge while significantly reducing model size and computational cost. By aligning the smaller model's predictions with a structured knowledge graph and ranked evidence, $\texttt{DRAG}$ effectively mitigates hallucinations and improves factual accuracy. We further present a case demonstrating how our framework mitigates user privacy risks and introduce a corresponding benchmark. Experimental evaluations on multiple benchmarks demonstrate that our method outperforms the prior competitive RAG methods like MiniRAG for SLMs by up to 27.7% using the same models, preserving high-level efficiency and reliability. With $\texttt{DRAG}$, we provide a practical and resource-efficient roadmap to deploying enhanced retrieval and generation capabilities in small-sized LLMs.
title DRAG: Distilling RAG for SLMs from LLMs to Transfer Knowledge and Mitigate Hallucination via Evidence and Graph-based Distillation
topic Computation and Language
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2506.01954