Training Language Models to Generate Text with Citations via Fine-grained Rewards

Chengyu Huang, Zeqiu Wu, Yushi Hu, Wenya Wang


Abstract
While recent Large Language Models (LLMs) have proven useful in answering user queries, they are prone to hallucination, and their responses often lack credibility due to missing references to reliable sources. An intuitive solution to these issues would be to include in-text citations referring to external documents as evidence. While previous works have directly prompted LLMs to generate in-text citations, their performances are far from satisfactory, especially when it comes to smaller LLMs. In this work, we propose an effective training framework using fine-grained rewards to teach LLMs to generate highly supportive and relevant citations, while ensuring the correctness of their responses. We also conduct a systematic analysis of applying these fine-grained rewards to common LLM training strategies, demonstrating its advantage over conventional practices. We conduct extensive experiments on Question Answering (QA) datasets taken from the ALCE benchmark and validate the model’s generalizability using EXPERTQA. On LLaMA-2-7B, the incorporation of fine-grained rewards achieves the best performance among the baselines, even surpassing that of GPT-3.5-turbo.
Anthology ID:
2024.acl-long.161
Volume:
Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
Month:
August
Year:
2024
Address:
Bangkok, Thailand
Editors:
Lun-Wei Ku, Andre Martins, Vivek Srikumar
Venue:
ACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
2926–2949
Language:
URL:
https://aclanthology.org/2024.acl-long.161
DOI:
10.18653/v1/2024.acl-long.161
Bibkey:
Cite (ACL):
Chengyu Huang, Zeqiu Wu, Yushi Hu, and Wenya Wang. 2024. Training Language Models to Generate Text with Citations via Fine-grained Rewards. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 2926–2949, Bangkok, Thailand. Association for Computational Linguistics.
Cite (Informal):
Training Language Models to Generate Text with Citations via Fine-grained Rewards (Huang et al., ACL 2024)
Copy Citation:
PDF:
https://aclanthology.org/2024.acl-long.161.pdf