arXiv 2606.174412026-08-12cs.HCcs.AIcs.CY

具有个性的患者：通过受控多样性与选择性披露实现逼真的患者模拟

Patients With Personality: Realistic Patient Simulation through Controlled Diversity and Selective Disclosure

Technical University of Munich（慕尼黑技术大学）；Munich Center for Machine Learning（慕尼黑机器学习中心）；TUM University Hospital（慕尼黑技术大学医院）；Google DeepMind（谷歌DeepMind）；Google Research（谷歌研究）；Imperial College London（伦敦帝国学院）

Moritz Schlager, Friederike Jungmann, Samuel Schmidgall, Philipp Raffler, Franziska Hartl, Eva Wende, Paula Roßmüller, Conrad Ketzer, Avinatan Hassidim, Dale R.… 展开作者

Moritz Schlager, Friederike Jungmann, Samuel Schmidgall, Philipp Raffler, Franziska Hartl, Eva Wende, Paula Roßmüller, Conrad Ketzer, Avinatan Hassidim, Dale R. Webster, Yossi Matias, Yun Liu, Daniel Rueckert, Mike Schaekermann, Paul Hager

备注：: 22 pages, 11 figures

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

提出PatientsWithPersonality框架，通过HEXACO人格参数化控制患者对话风格、合作性和信息披露，生成逼真且多样化的虚拟患者，在临床评估中接近真实演员表现。

摘要：

模拟逼真的患者交互是在没有耗时且昂贵的用户研究的情况下大规模测试LLMs临床应用的关键要求。然而，现有方法通常缺乏真实性和可控性，常常在未提示的情况下过度分享信息，并且未能捕捉患者行为的广泛变异性。在这里，我们引入了PatientsWithPersonality (PWP)，一个患者模拟框架，通过在潜在患者状态上显式的人格参数化生成逼真且多样化的虚拟患者响应。基于HEXACO（一个用于量化和参数化人类行为特征的六维人格空间），我们的方法能够在统一框架内对对话风格、合作性和信息披露进行细粒度控制。在临床医生评估中，PWP被认为几乎与记录的人类演员一样逼真，并且明显优于先前的模拟器，同时被标记为“信息过多”的频率远低于前者。基于HEXACO轴的条件化产生的人格特质可由临床医生和自动评估者恢复，其行为足迹比最接近的基线宽得多，并防止过度分享。总之，我们的框架通过逼真且可操控的患者模拟器，为更准确且信息丰富的LLM基准测试铺平了道路。

英文摘要：

Simulating realistic patient interactions is a key requirement to testing clinical applications of LLMs at scale without time-consuming and expensive user studies. However, existing approaches often lack realism and controllability, often oversharing information unprompted, and failing to capture the wide variability of patient behavior. Here, we introduce PatientsWithPersonality (PWP), a patient simulation framework that generates realistic yet diverse virtual patient responses through explicit personality parametrization over a latent patient state. Grounded in HEXACO, a six-dimensional personality space used to quantify and parameterize human behavioral traits, our approach enables fine-grained control over conversational style, cooperativeness, and information disclosure within a unified framework. In a clinician evaluation, PWP is judged nearly as realistic as recorded human actors and clearly ahead of prior simulators, while being flagged as "too informative" far less often. Conditioning on HEXACO axes yields personas whose configured traits are recoverable by both clinicians and an autorater, span a substantially wider behavioral footprint than the closest baseline, and prevent oversharing. Altogether, our framework paves the way for more accurate and informative LLM benchmarking through our realistic and steerable patient simulator.

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。