EMPOWER: Evolutionary Medical Prompt Optimization With Reinforcement Learning

TL;DR

EMPOWER is an evolutionary framework that improves medical prompts for LLMs by incorporating medical knowledge, safety, and clinical reasoning, leading to more accurate and domain-specific outputs in healthcare applications.

Contribution

It introduces a novel medical prompt optimization method combining specialized representation learning, multi-dimensional evaluation, and structure-preserving algorithms.

Findings

24.7% reduction in factually incorrect content

19.6% improvement in domain specificity

15.3% higher clinician preference

Abstract

Prompt engineering significantly influences the reliability and clinical utility of Large Language Models (LLMs) in medical applications. Current optimization approaches inadequately address domain-specific medical knowledge and safety requirements. This paper introduces EMPOWER, a novel evolutionary framework that enhances medical prompt quality through specialized representation learning, multi-dimensional evaluation, and structure-preserving algorithms. Our methodology incorporates: (1) a medical terminology attention mechanism, (2) a comprehensive assessment architecture evaluating clarity, specificity, clinical relevance, and factual accuracy, (3) a component-level evolutionary algorithm preserving clinical reasoning integrity, and (4) a semantic verification module ensuring adherence to medical knowledge. Evaluation across diagnostic, therapeutic, and educational tasks demonstrates significant improvements: 24.7% reduction in factually incorrect content, 19.6% enhancement in domain specificity, and 15.3% higher clinician preference in blinded evaluations. The framework addresses critical challenges in developing clinically appropriate prompts, facilitating more responsible integration of LLMs into healthcare settings.