Grammar-Guided Evolutionary Search for Discrete Prompt Optimisation

TL;DR

This paper introduces a grammar-guided evolutionary search method for discrete prompt optimization that outperforms existing approaches on smaller language models and complex tasks by synthesizing and fine-tuning prompt programs.

Contribution

It presents a novel two-phase evolutionary approach combining grammar-guided genetic programming and local search for improved prompt optimization.

Findings

Outperforms state-of-the-art prompt optimization methods on small LLMs.

Effective in complex, domain-specific tasks with minimal performance degradation.

Demonstrates robustness across multiple models and tasks.

Abstract

Prompt engineering has proven to be a crucial step in leveraging pretrained large language models (LLMs) in solving various real-world tasks. Numerous solutions have been proposed that seek to automate prompt engineering by using the model itself to edit prompts. However, the majority of state-of-the-art approaches are evaluated on tasks that require minimal prompt templates and on very large and highly capable LLMs. In contrast, solving complex tasks that require detailed information to be included in the prompt increases the amount of text that needs to be optimised. Furthermore, smaller models have been shown to be more sensitive to prompt design. To address these challenges, we propose an evolutionary search approach to automated discrete prompt optimisation consisting of two phases. In the first phase, grammar-guided genetic programming is invoked to synthesise prompt-creating programmes by searching the space of programmes populated by function compositions of syntactic, dictionary-based and LLM-based prompt-editing functions. In the second phase, local search is applied to explore the neighbourhoods of best-performing programmes in an attempt to further fine-tune their performance. Our approach outperforms three state-of-the-art prompt optimisation approaches, PromptWizard, OPRO, and RL-Prompt, on three relatively small general-purpose LLMs in four domain-specific challenging tasks. We also illustrate several examples where these benchmark methods suffer relatively severe performance degradation, while our approach improves performance in almost all task-model combinations, only incurring minimal degradation when it does not.