NICE: To Optimize In-Context Examples or Not?

Pragya Srivastava, Satvik Golechha, Amit Deshpande, Amit Sharma


Abstract
Recent work shows that in-context learning and optimization of in-context examples (ICE) can significantly improve the accuracy of large language models (LLMs) on a wide range of tasks, leading to an apparent consensus that ICE optimization is crucial for better performance. However, most of these studies assume a fixed or no instruction provided in the prompt. We challenge this consensus by investigating the necessity of optimizing ICE when task-specific instructions are provided and find that there are many tasks for which it yields diminishing returns. In particular, using a diverse set of tasks and a systematically created instruction set with gradually added details, we find that as the prompt instruction becomes more detailed, the returns on ICE optimization diminish. To characterize this behavior, we introduce a task-specific metric called Normalized Invariability to Choice of Examples (NICE) that quantifies the learnability of tasks from a given instruction, and provides a heuristic to help decide whether to optimize instructions or ICE for a new task. Given a task, the proposed metric can reliably predict the utility of optimizing ICE compared to using random ICE. Our code is available at [https://github.com/microsoft/nice-icl](https://github.com/microsoft/nice-icl).
Anthology ID:
2024.acl-long.300
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:
5494–5510
Language:
URL:
https://aclanthology.org/2024.acl-long.300
DOI:
10.18653/v1/2024.acl-long.300
Bibkey:
Cite (ACL):
Pragya Srivastava, Satvik Golechha, Amit Deshpande, and Amit Sharma. 2024. NICE: To Optimize In-Context Examples or Not?. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 5494–5510, Bangkok, Thailand. Association for Computational Linguistics.
Cite (Informal):
NICE: To Optimize In-Context Examples or Not? (Srivastava et al., ACL 2024)
Copy Citation:
PDF:
https://preview.aclanthology.org/autopr/2024.acl-long.300.pdf