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

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Liu, Dou, Long, Ying, Zuoqiu, Sophia, Xie, Kaipeng, Yang, Runze, Liu, Di, Li, Kang, Lin, Yiting, Liu, Hanyi, Yin, Rong, Tang, Tian
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866917098734747648
author Liu, Dou
Long, Ying
Zuoqiu, Sophia
Xie, Kaipeng
Yang, Runze
Liu, Di
Li, Kang
Lin, Yiting
Liu, Hanyi
Yin, Rong
Tang, Tian
author_facet Liu, Dou
Long, Ying
Zuoqiu, Sophia
Xie, Kaipeng
Yang, Runze
Liu, Di
Li, Kang
Lin, Yiting
Liu, Hanyi
Yin, Rong
Tang, Tian
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2511_18084
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle The Alignment Paradox of Medical Large Language Models in Infertility Care: Decoupling Algorithmic Improvement from Clinical Decision-making Quality
Liu, Dou
Long, Ying
Zuoqiu, Sophia
Xie, Kaipeng
Yang, Runze
Liu, Di
Li, Kang
Lin, Yiting
Liu, Hanyi
Yin, Rong
Tang, Tian
Machine Learning
Artificial Intelligence
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.
title The Alignment Paradox of Medical Large Language Models in Infertility Care: Decoupling Algorithmic Improvement from Clinical Decision-making Quality
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2511.18084