TY - GEN
T1 - Can LLMs Learn Uncertainty on Their Own? Expressing Uncertainty Effectively in A Self-Training Manner
AU - Liu, Shudong
AU - Li, Zhaocong
AU - Liu, Xuebo
AU - Zhan, Runzhe
AU - Wong, Derek F.
AU - Chao, Lidia S.
AU - Zhang, Min
N1 - Publisher Copyright:
© 2024 Association for Computational Linguistics.
PY - 2024
Y1 - 2024
N2 - Large language models (LLMs) often exhibit excessive, random, and uninformative uncertainty, rendering them unsuitable for decision-making in human-computer interactions. In this paper, we aim to instigate a heightened awareness of self-uncertainty in LLMs, enabling them to express uncertainty more effectively. To accomplish this, we propose an uncertainty-aware instruction tuning (UaIT) method, aligning LLMs' perception with the probabilistic uncertainty of the generation. We conducted experiments using LLaMA2 and Mistral on multiple free-form QA tasks. Experimental results revealed a surprising 45.2% improvement in the effectiveness of uncertainty expression by LLMs, accompanied by reasonably good out-of-domain generalization capabilities. Moreover, this uncertainty expression can serve as a valuable real-time basis for human decision-making, e.g., retrieving external documents and incorporating stronger LLMs.
AB - Large language models (LLMs) often exhibit excessive, random, and uninformative uncertainty, rendering them unsuitable for decision-making in human-computer interactions. In this paper, we aim to instigate a heightened awareness of self-uncertainty in LLMs, enabling them to express uncertainty more effectively. To accomplish this, we propose an uncertainty-aware instruction tuning (UaIT) method, aligning LLMs' perception with the probabilistic uncertainty of the generation. We conducted experiments using LLaMA2 and Mistral on multiple free-form QA tasks. Experimental results revealed a surprising 45.2% improvement in the effectiveness of uncertainty expression by LLMs, accompanied by reasonably good out-of-domain generalization capabilities. Moreover, this uncertainty expression can serve as a valuable real-time basis for human decision-making, e.g., retrieving external documents and incorporating stronger LLMs.
UR - https://www.scopus.com/pages/publications/85212211161
U2 - 10.18653/v1/2024.emnlp-main.1205
DO - 10.18653/v1/2024.emnlp-main.1205
M3 - 会议稿件
AN - SCOPUS:85212211161
T3 - EMNLP 2024 - 2024 Conference on Empirical Methods in Natural Language Processing, Proceedings of the Conference
SP - 21635
EP - 21645
BT - EMNLP 2024 - 2024 Conference on Empirical Methods in Natural Language Processing, Proceedings of the Conference
A2 - Al-Onaizan, Yaser
A2 - Bansal, Mohit
A2 - Chen, Yun-Nung
PB - Association for Computational Linguistics (ACL)
T2 - 2024 Conference on Empirical Methods in Natural Language Processing, EMNLP 2024
Y2 - 12 November 2024 through 16 November 2024
ER -