Saved in:
Bibliographic Details
Main Authors: Chen, Yinda, He, Yangfan, Yang, Jing, Zhang, Dapeng, Yuan, Zhenlong, Khan, Muhammad Attique, Baili, Jamel, Yee, Por Lip
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2508.17703
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912552067268608
author Chen, Yinda
He, Yangfan
Yang, Jing
Zhang, Dapeng
Yuan, Zhenlong
Khan, Muhammad Attique
Baili, Jamel
Yee, Por Lip
author_facet Chen, Yinda
He, Yangfan
Yang, Jing
Zhang, Dapeng
Yuan, Zhenlong
Khan, Muhammad Attique
Baili, Jamel
Yee, Por Lip
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2508_17703
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle EMPOWER: Evolutionary Medical Prompt Optimization With Reinforcement Learning
Chen, Yinda
He, Yangfan
Yang, Jing
Zhang, Dapeng
Yuan, Zhenlong
Khan, Muhammad Attique
Baili, Jamel
Yee, Por Lip
Computation and Language
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.
title EMPOWER: Evolutionary Medical Prompt Optimization With Reinforcement Learning
topic Computation and Language
url https://arxiv.org/abs/2508.17703