RAG-HAT: A Hallucination-Aware Tuning Pipeline for LLM in Retrieval-Augmented Generation

Juntong Song, Xingguang Wang, Juno Zhu, Yuanhao Wu, Xuxin Cheng, Randy Zhong, Cheng Niu


Abstract
Retrieval-augmented generation (RAG) has emerged as a significant advancement in the field of large language models (LLMs). By integrating up-to-date information not available during their initial training, RAG greatly enhances the practical utility of LLMs in real-world applications. However, even with RAG, LLMs can still produce inaccurate outputs, such as distorting or misinterpreting source content, posing risks in high-trust scenarios. To address these issues, we introduce a novel approach called Hallucination Aware Tuning (HAT). This method involves training hallucination detection models that generate detection labels and provide detailed descriptions of the detected hallucinations. Utilizing these detection results—particularly the hallucination descriptions—GPT-4 Turbo is employed to correct any detected hallucinations. The corrected outputs, free of hallucinations, along with the original versions, are used to create a preference dataset for Direct Preference Optimization (DPO) training. The fine-tuning through DPO leads to LLMs that exhibit a reduced rate of hallucinations and deliver improved answer quality.
Anthology ID:
2024.emnlp-industry.113
Volume:
Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing: Industry Track
Month:
November
Year:
2024
Address:
Miami, Florida, US
Editors:
Franck Dernoncourt, Daniel Preoţiuc-Pietro, Anastasia Shimorina
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
1548–1558
Language:
URL:
https://aclanthology.org/2024.emnlp-industry.113
DOI:
Bibkey:
Cite (ACL):
Juntong Song, Xingguang Wang, Juno Zhu, Yuanhao Wu, Xuxin Cheng, Randy Zhong, and Cheng Niu. 2024. RAG-HAT: A Hallucination-Aware Tuning Pipeline for LLM in Retrieval-Augmented Generation. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing: Industry Track, pages 1548–1558, Miami, Florida, US. Association for Computational Linguistics.
Cite (Informal):
RAG-HAT: A Hallucination-Aware Tuning Pipeline for LLM in Retrieval-Augmented Generation (Song et al., EMNLP 2024)
Copy Citation:
PDF:
https://aclanthology.org/2024.emnlp-industry.113.pdf
Poster:
 2024.emnlp-industry.113.poster.pdf
Presentation:
 2024.emnlp-industry.113.presentation.pdf