Better Zero-Shot Reasoning with Self-Adaptive Prompting

TL;DR

This paper introduces COSP, a self-adaptive prompting method that enhances zero-shot reasoning in large language models by automatically selecting examples from their own outputs, improving performance without human-crafted data.

Contribution

COSP is a novel prompt design technique that automatically constructs effective examples from LLM outputs, eliminating the need for handcrafted responses or labels.

Findings

COSP improves zero-shot performance by up to 15%.

COSP matches or exceeds few-shot baselines across various tasks.

COSP reduces human effort in prompt engineering.

Abstract

Modern large language models (LLMs) have demonstrated impressive capabilities at sophisticated tasks, often through step-by-step reasoning similar to humans. This is made possible by their strong few and zero-shot abilities -- they can effectively learn from a handful of handcrafted, completed responses ("in-context examples"), or are prompted to reason spontaneously through specially designed triggers. Nonetheless, some limitations have been observed. First, performance in the few-shot setting is sensitive to the choice of examples, whose design requires significant human effort. Moreover, given the diverse downstream tasks of LLMs, it may be difficult or laborious to handcraft per-task labels. Second, while the zero-shot setting does not require handcrafting, its performance is limited due to the lack of guidance to the LLMs. To address these limitations, we propose Consistency-based Self-adaptive Prompting (COSP), a novel prompt design method for LLMs. Requiring neither handcrafted responses nor ground-truth labels, COSP selects and builds the set of examples from the LLM zero-shot outputs via carefully designed criteria that combine consistency, diversity and repetition. In the zero-shot setting for three different LLMs, we show that using only LLM predictions, COSP improves performance up to 15% compared to zero-shot baselines and matches or exceeds few-shot baselines for a range of reasoning tasks.