Debiasing In-Context Learning by Instructing LLMs How to Follow Demonstrations

Lvxue Li, Jiaqi Chen, Xinyu Lu, Yaojie Lu, Hongyu Lin, Shuheng Zhou, Huijia Zhu, Weiqiang Wang, Zhongyi Liu, Xianpei Han, Le Sun


Abstract
In-context learning(ICL) has gained considerable attention due to its data efficiency and task adaptability. Unfortunately, ICL suffers from the demonstration bias, i.e., its performance and robustness are severely affected by the selection and ordering of demonstrations. In this paper, we identify that such demonstration bias may primarily stem from the semantic ambiguity induced by demonstrations, i.e., a demonstration may indicate multiple input-to-label mappings and its mapping can be interpreted differently in different contexts by LLMs. Such semantic ambiguity disrupts task comprehension during ICL and results in performance fluctuations. To resolve the semantic ambiguity problem, this paper further proposes two de-biasing strategies to mitigate demonstration bias in in-context learning. Experiments on six datasets show that our methods can effectively alleviate demonstration bias and significantly improve task performance.
Anthology ID:
2024.findings-acl.430
Volume:
Findings of the Association for Computational Linguistics ACL 2024
Month:
August
Year:
2024
Address:
Bangkok, Thailand and virtual meeting
Editors:
Lun-Wei Ku, Andre Martins, Vivek Srikumar
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
7203–7215
Language:
URL:
https://aclanthology.org/2024.findings-acl.430
DOI:
Bibkey:
Cite (ACL):
Lvxue Li, Jiaqi Chen, Xinyu Lu, Yaojie Lu, Hongyu Lin, Shuheng Zhou, Huijia Zhu, Weiqiang Wang, Zhongyi Liu, Xianpei Han, and Le Sun. 2024. Debiasing In-Context Learning by Instructing LLMs How to Follow Demonstrations. In Findings of the Association for Computational Linguistics ACL 2024, pages 7203–7215, Bangkok, Thailand and virtual meeting. Association for Computational Linguistics.
Cite (Informal):
Debiasing In-Context Learning by Instructing LLMs How to Follow Demonstrations (Li et al., Findings 2024)
Copy Citation:
PDF:
https://preview.aclanthology.org/nschneid-patch-4/2024.findings-acl.430.pdf