Universal Self-Adaptive Prompting

TL;DR

Universal Self-Adaptive Prompting (USP) automatically designs prompts for zero-shot learning in large language models by categorizing tasks and selecting optimal queries, significantly improving performance across diverse NLP tasks.

Contribution

USP introduces a fully automated, versatile prompt design method that adapts to various NLP tasks using minimal unlabeled data and a task categorization approach.

Findings

USP outperforms standard zero-shot baselines.

USP matches or exceeds few-shot performance.

Effective across 40+ NLP tasks.

Abstract

A hallmark of modern large language models (LLMs) is their impressive general zero-shot and few-shot abilities, often elicited through in-context learning (ICL) via prompting. However, while highly coveted and being the most general, zero-shot performances in LLMs are still typically weaker due to the lack of guidance and the difficulty of applying existing automatic prompt design methods in general tasks when ground-truth labels are unavailable. In this study, we address this by presenting Universal Self-Adaptive Prompting (USP), an automatic prompt design approach specifically tailored for zero-shot learning (while compatible with few-shot). Requiring only a small amount of unlabeled data and an inference-only LLM, USP is highly versatile: to achieve universal prompting, USP categorizes a possible NLP task into one of the three possible task types and then uses a corresponding selector to select the most suitable queries and zero-shot model-generated responses as pseudo-demonstrations, thereby generalizing ICL to the zero-shot setup in a fully automated way. We evaluate USP with PaLM and PaLM 2 models and demonstrate performances that are considerably stronger than standard zero-shot baselines and often comparable to or even superior to few-shot baselines across more than 40 natural language understanding, natural language generation, and reasoning tasks.