Learning New Tasks from a Few Examples with Soft-Label Prototypes

TL;DR

This paper introduces a novel few-shot learning method using soft-label prototypes that effectively captures class distributions, enabling superior performance on unseen NLP tasks with minimal examples and efficient parameter use.

Contribution

The proposed soft-label prototypes approach is a new method for few-shot NLP learning that outperforms existing models and can be integrated into meta-learning frameworks.

Findings

Achieves superior performance on most tested tasks with very few examples.

Highly parameter efficient compared to existing methods.

Can be integrated into meta-learning for improved results.

Abstract

Existing approaches to few-shot learning in NLP rely on large language models (LLMs) and/or fine-tuning of these to generalise on out-of-distribution data. In this work, we propose a novel few-shot learning approach based on soft-label prototypes (SLPs) designed to collectively capture the distribution of different classes across the input domain space. We focus on learning previously unseen NLP tasks from very few examples (4, 8, 16) per class and experimentally demonstrate that our approach achieves superior performance on the majority of tested tasks in this data-lean setting while being highly parameter efficient. We also show that our few-shot adaptation method can be integrated into more generalised learning settings, primarily meta-learning, to yield superior performance against strong baselines.