Skill-Based Few-Shot Selection for In-Context Learning

TL;DR

This paper introduces Skill-KNN, a skill-based few-shot selection method for in-context learning that improves example selection by focusing on task-relevant skills, without requiring model training or fine-tuning.

Contribution

Skill-KNN is a novel, training-free approach that enhances in-context learning by selecting examples based on skill descriptions, reducing bias from surface features.

Findings

Outperforms existing methods across multiple datasets

Effective across diverse backbone models

Addresses bias from surface language features

Abstract

In-context learning is the paradigm that adapts large language models to downstream tasks by providing a few examples. Few-shot selection -- selecting appropriate examples for each test instance separately -- is important for in-context learning. In this paper, we propose Skill-KNN, a skill-based few-shot selection method for in-context learning. The key advantages of Skill-KNN include: (1) it addresses the problem that existing methods based on pre-trained embeddings can be easily biased by surface natural language features that are not important for the target task; (2) it does not require training or fine-tuning of any models, making it suitable for frequently expanding or changing example banks. The key insight is to optimize the inputs fed into the embedding model, rather than tuning the model itself. Technically, Skill-KNN generates the skill-based descriptions for each test case and candidate example by utilizing a pre-processing few-shot prompting, thus eliminating unimportant surface features. Experimental results across five cross-domain semantic parsing datasets and six backbone models show that Skill-KNN significantly outperforms existing methods.