Progressive-Hint Prompting Improves Reasoning in Large Language Models

TL;DR

Progressive-Hint Prompting (PHP) is a new method that enhances reasoning in large language models by using previous answers as hints, significantly improving accuracy and efficiency across multiple benchmarks.

Contribution

The paper introduces PHP, a novel prompting technique that leverages iterative hints from previous answers, complementing existing methods like CoT and self-consistency to boost reasoning performance.

Findings

PHP improves GSM8K accuracy by 4.2% with greedy decoding.

PHP reduces sample paths by 46.17% with self-consistency.

PHP achieves state-of-the-art results on SVAMP, GSM8K, AQuA, and MATH.

Abstract

The performance of Large Language Models (LLMs) in reasoning tasks depends heavily on prompt design, with Chain-of-Thought (CoT) and self-consistency being critical methods that enhance this ability. However, these methods do not fully exploit the answers generated by the LLM to guide subsequent responses. This paper proposes a new prompting method, named Progressive-Hint Prompting (PHP), that enables automatic multiple interactions between users and LLMs by using previously generated answers as hints to progressively guide toward the correct answers. PHP is orthogonal to CoT and self-consistency, making it easy to combine with state-of-the-art techniques to further improve performance. We conducted extensive and comprehensive experiments on seven benchmarks. The results show that PHP significantly improves accuracy while remaining highly efficient. For instance, with text-davinci-003, we observed a 4.2% improvement on GSM8K with greedy decoding compared to Complex CoT, and a 46.17% reduction in sample paths with self-consistency. With GPT-4 and PHP, we achieve state-of-the-art performances on SVAMP (89.1% -> 91.9%), GSM8K (92% -> 95.5%), AQuA (76.4% -> 79.9%) and MATH (50.3% -> 53.9%).