Addressing Order Sensitivity of In-Context Demonstration Examples in Causal Language Models

Yanzheng Xiang, Hanqi Yan, Lin Gui, Yulan He


Abstract
In-context learning has become a popular paradigm in natural language processing. However, its performance can be significantly influenced by the order of in-context demonstration examples. In this paper, we found that causal language models (CausalLMs) are more sensitive to this order compared to prefix language models (PrefixLMs). We attribute this phenomenon to the auto-regressive attention masks within CausalLMs, which restrict each token from accessing information from subsequent tokens. This results in different receptive fields for samples at different positions, thereby leading to representation disparities across positions. To tackle this challenge, we introduce an unsupervised fine-tuning method, termed the Information-Augmented and Consistency-Enhanced approach. This approach utilizes contrastive learning to align representations of in-context examples across different positions and introduces a consistency loss to ensure similar representations for inputs with different permutations. This enhances the model’s predictive consistency across permutations. Experimental results on five benchmarks suggest that our proposed method can reduce the sensitivity of CausalLMs to the order of in-context examples and exhibit robust generalizability, particularly when demonstrations are sourced from a candidate pool different from that used in the training phase, or when the number of in-context examples differs from what is used during training.
Anthology ID:
2024.findings-acl.386
Volume:
Findings of the Association for Computational Linguistics: ACL 2024
Month:
August
Year:
2024
Address:
Bangkok, Thailand
Editors:
Lun-Wei Ku, Andre Martins, Vivek Srikumar
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
6467–6481
Language:
URL:
https://aclanthology.org/2024.findings-acl.386
DOI:
10.18653/v1/2024.findings-acl.386
Bibkey:
Cite (ACL):
Yanzheng Xiang, Hanqi Yan, Lin Gui, and Yulan He. 2024. Addressing Order Sensitivity of In-Context Demonstration Examples in Causal Language Models. In Findings of the Association for Computational Linguistics: ACL 2024, pages 6467–6481, Bangkok, Thailand. Association for Computational Linguistics.
Cite (Informal):
Addressing Order Sensitivity of In-Context Demonstration Examples in Causal Language Models (Xiang et al., Findings 2024)
Copy Citation:
PDF:
https://aclanthology.org/2024.findings-acl.386.pdf