Evolutionary Pre-Prompt Optimization for Mathematical Reasoning

TL;DR

This paper introduces Evolutionary Pre-Prompt Optimization (EPPO), an evolutionary algorithm-based method that improves the selection of few-shot exemplars for chain-of-thought prompting, significantly boosting mathematical reasoning accuracy in large language models.

Contribution

It presents a novel evolutionary optimization approach for selecting effective few-shot prompts, outperforming naive methods and enhancing reasoning performance in LLMs.

Findings

EPPO exceeds 10 points in exact match scores on GSM8k and MathQA datasets.

EPPO's improvements are consistent across different contexts.

Combining EPPO with self-consistency further amplifies performance gains.

Abstract

Recent advancements have highlighted that large language models (LLMs), when given a small set of task-specific examples, demonstrate remarkable proficiency, a capability that extends to complex reasoning tasks. In particular, the combination of few-shot learning with the chain-of-thought (CoT) approach has been pivotal in steering models towards more logically consistent conclusions [Wei et al. 2022b]. This paper explores the optimization of example selection for designing effective CoT pre-prompts and shows that the choice of the optimization algorithm, typically in favor of comparison-based methods such as evolutionary computation, significantly enhances efficacy and feasibility. Specifically, thanks to a limited exploitative and overfitted optimization, Evolutionary Pre-Prompt Optimization (EPPO) brings an improvement over the naive few-shot approach, exceeding 10 absolute points in exact match scores on benchmark datasets such as GSM8k and MathQA. These gains are consistent across various contexts and are further amplified when integrated with self-consistency (SC).