Stepwise Perplexity-Guided Refinement for Efficient Chain-of-Thought Reasoning in Large Language Models

TL;DR

This paper introduces a perplexity-guided method to identify and focus on critical reasoning steps in Chain-of-Thought prompting, improving efficiency without sacrificing accuracy in large language models.

Contribution

It presents a novel stepwise refinement approach that uses perplexity to select essential reasoning steps, reducing computational costs in CoT reasoning.

Findings

Improved reasoning efficiency while maintaining accuracy

Effective identification of critical reasoning steps using perplexity

Enhanced balance between speed and performance in LLMs

Abstract

Chain-of-Thought (CoT) reasoning, which breaks down complex tasks into intermediate reasoning steps, has significantly enhanced the performance of large language models (LLMs) on challenging tasks. However, the detailed reasoning process in CoT often incurs long generation times and high computational costs, partly due to the inclusion of unnecessary steps. To address this, we propose a method to identify critical reasoning steps using perplexity as a measure of their importance: a step is deemed critical if its removal causes a significant increase in perplexity. Our method enables models to focus solely on generating these critical steps. This can be achieved through two approaches: refining demonstration examples in few-shot CoT or fine-tuning the model using selected examples that include only critical steps. Comprehensive experiments validate the effectiveness of our method, which achieves a better balance between the reasoning accuracy and efficiency of CoT.