A Human-Centric Pipeline for Aligning Large Language Models with Chinese Medical Ethics
Haoan Jin, Han Ying, Jiacheng Ji, Hanhui Xu, Mengyue Wu
TL;DR
This work addresses the challenge of aligning large language models with medical ethics in China, where high-stakes clinical decisions demand strict normative compliance. It introduces MedES, a scenario-centric benchmark derived from 260 regulatory and ethical sources to cover 12 high-risk clinical scenarios, and a guardian-in-the-loop pipeline that uses a trained evaluator to guide supervised fine-tuning and iterative alignment. The approach yields measurable gains, with a 7B parameter model outperforming a substantially larger baseline on core ethical tasks through multi-round, evaluator-guided refinement. The framework demonstrates practical viability for ethical alignment in medical AI within the Chinese healthcare domain and suggests modular adaptation to other legal/cultural settings via changes to the normative corpus. Overall, the paper highlights the importance of fine-grained, dynamic ethical supervision and iterative data-driven optimization for safe clinical AI deployment.
Abstract
Recent advances in large language models have enabled their application to a range of healthcare tasks. However, aligning LLMs with the nuanced demands of medical ethics, especially under complex real world scenarios, remains underexplored. In this work, we present MedES, a dynamic, scenario-centric benchmark specifically constructed from 260 authoritative Chinese medical, ethical, and legal sources to reflect the challenges in clinical decision-making. To facilitate model alignment, we introduce a guardian-in-the-loop framework that leverages a dedicated automated evaluator (trained on expert-labeled data and achieving over 97% accuracy within our domain) to generate targeted prompts and provide structured ethical feedback. Using this pipeline, we align a 7B-parameter LLM through supervised fine-tuning and domain-specific preference optimization. Experimental results, conducted entirely within the Chinese medical ethics context, demonstrate that our aligned model outperforms notably larger baselines on core ethical tasks, with observed improvements in both quality and composite evaluation metrics. Our work offers a practical and adaptable framework for aligning LLMs with medical ethics in the Chinese healthcare domain, and suggests that similar alignment pipelines may be instantiated in other legal and cultural environments through modular replacement of the underlying normative corpus.