arXiv 2511.180842025-11-22cs.LGcs.AI

医学大语言模型在不孕症护理中的对齐悖论：解耦算法改进与临床决策质量

The Alignment Paradox of Medical Large Language Models in Infertility Care: Decoupling Algorithmic Improvement from Clinical Decision-making Quality

Department of Obstetrics and Gynecology, West China Second University Hospital（妇产科部门，西昌第二大学医院）；Reproductive Medical Center, Department of Obstetrics and Gynecology, West China Second University Hospital（生殖医学中心，妇产科部门，西昌第二大学医院）

Dou Liu, Ying Long, Sophia Zuoqiu, Kaipeng Xie, Runze Yang, Di Liu, Kang Li, Yiting Lin, Hanyi Liu, Rong Yin, Tian Tang

备注：: 22 pages 5 figures

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本研究发现，尽管GRPO在算法准确性上表现最佳，但临床医生更偏好SFT模型，因其推理更清晰且治疗可行性更高，揭示了算法改进与临床信任之间的对齐悖论。

摘要：

大型语言模型（LLMs）正在越来越多地被应用于临床决策支持，但将其与现实医学多方面的推理路径对齐仍是一个重大挑战。使用超过8000份不孕治疗记录，我们系统地评估了四种对齐策略：监督微调（SFT）、直接偏好优化（DPO）、组相对策略优化（GRPO）和上下文学习（ICL），通过结合自动基准测试和盲审医生在环评估的双层框架进行评估。GRPO在多个决策层上实现了最高的算法准确性，证实了基于强化学习的优化在结构化预测任务中的价值。然而，临床医生始终偏好SFT模型， citing更清晰的推理过程（p=0.035）和更高的治疗可行性（p=0.019）。在盲测成对比较中，SFT获得最高获胜率（51.2%），优于GRPO（26.2%）甚至医生的原始决策（22.7%）。这些结果揭示了一个对齐悖论：算法改进不一定能转化为更高的临床信任，可能与以人类为中心的偏好相背离。我们的发现强调了需要优先考虑临床可解释和实际可行推理的对齐策略，而不是仅仅优化决策层面的准确性。

英文摘要：

Large language models (LLMs) are increasingly adopted in clinical decision support, yet aligning them with the multifaceted reasoning pathways of real-world medicine remains a major challenge. Using more than 8,000 infertility treatment records, we systematically evaluate four alignment strategies: Supervised Fine-Tuning (SFT), Direct Preference Optimization (DPO), Group Relative Policy Optimization (GRPO), and In-Context Learning (ICL) through a dual-layer framework combining automatic benchmarks with blinded doctor-in-the-loop assessments. GRPO achieves the highest algorithmic accuracy across multiple decision layers, confirming the value of reinforcement-based optimization for structured prediction tasks. However, clinicians consistently prefer the SFT model, citing clearer reasoning processes (p = 0.035) and higher therapeutic feasibility (p = 0.019). In blinded pairwise comparisons, SFT attains the highest winning rate (51.2%), outperforming both GRPO (26.2%) and even physicians' original decisions (22.7%). These results reveal an alignment paradox: algorithmic improvements do not necessarily translate into higher clinical trust, and may diverge from human-centered preferences. Our findings highlight the need for alignment strategies that prioritize clinically interpretable and practically feasible reasoning, rather than solely optimizing decision-level accuracy.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。