Flatness-Aware Prompt Selection Improves Accuracy and Sample Efficiency

With growing capabilities of large language models, prompting them has become the dominant way to access them. This has motivated the development of strategies for automatically selecting effective language prompts. In this paper, we introduce prompt flatness, a new metric to quantify the expected u...

Full description

Saved in:
Bibliographic Details
Published inarXiv.org
Main Authors Shen, Lingfeng, Tan, Weiting, Zheng, Boyuan, Khashabi, Daniel
Format Paper
LanguageEnglish
Published Ithaca Cornell University Library, arXiv.org 23.10.2023
Subjects
Accuracy
Flatness
Large language models
Perturbation
Regularization
Online AccessGet full text

Cover

More Information
Summary:With growing capabilities of large language models, prompting them has become the dominant way to access them. This has motivated the development of strategies for automatically selecting effective language prompts. In this paper, we introduce prompt flatness, a new metric to quantify the expected utility of a language prompt. This metric is inspired by flatness regularization in statistical learning that quantifies the robustness of the model towards its parameter perturbations. We provide theoretical foundations for this metric and its relationship with other prompt selection metrics, providing a comprehensive understanding of existing methods. Empirically, we show that combining prompt flatness with existing metrics improves both performance and sample efficiency. Our metric outperforms the previous prompt selection metrics with an average increase of 5% in accuracy and 10% in Pearson correlation across 6 classification benchmarks.
ISSN:2331-8422