arXiv cs.CL· Jiawei Zhang, Huachuan Qiu, Ruocheng Guo, Zijun Shi·· 11 小时前AI 评分31
PsyInterviewee：心理健康访谈模拟中的信息不变性与关系适应性
PsyInterviewee: Informational Invariance and Relational Adaptivity in Mental Health Interview Simulation
AI 导读
PsyInterviewee 将受保护的真相通道与动态交互通道分离，并通过非对称写权限让对话上下文只更新披露与行为状态、不授权新增临床事实，从而在保持画像临床事实不变的同时适配访谈互动。在基于 LLM 访谈者的自动评测中，该框架在两条轴上的多数指标均优于四个代表性基线，并由两个 LLM 评委打分；消融实验显示两个通道分别贡献临床事实锚定与回应调节。
正文 · 原文
Abstract:LLM-based simulated interviewees support interviewer training and the evaluation of mental health interviewing systems. Existing studies have primarily focused on improving human-likeness through richer patient profiles and behavioral modeling. However, human-likeness alone does not establish their reliability. Reliable simulation requires preserving profile-grounded clinical facts (informational invariance) while adapting disclosure and interaction behavior to the interview (relational adaptivity). To this end, we introduce PsyInterviewee, which separates a protected truth channel from a dynamic interaction channel and verifies generated clinical claims against the profile. Specifically, asymmetric write permissions allow dialogue context to update disclosure and behavioral states without authorizing new clinical facts; these states guide what profile-supported information is revealed and how it is expressed. In an automated evaluation of interactions between LLM-based interviewers and interviewees, PsyInterviewee achieves higher scores than four representative baselines on most metrics across both axes, as assessed by two LLM judges. Ablations suggest that the two channels contribute to clinical grounding and response modulation, respectively.
| Comments: | 38 pages, including appendices. Code: this https URL |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.41289 [cs.CL] |
| (or arXiv:2610.41289v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.41289 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Jiawei Zhang [view email]
[v1]
Fri, 30 Oct 2026 05:50:25 UTC (1,266 KB)
来源：arXiv cs.CL · arxiv.org