Angela Chen, Siwei Jin, Catherine Bao, Canwen Wang, Robert E. Kraut, Tongshuang Wu, Haiyi Zhu — cs.CY, cs.HC 📄 Full text (arXiv)
The Adaptive Virtual Patient (AVP) is an LLM-driven simulated patient for psychotherapy training that adapts its disclosure behavior — from guarded through moderate openness to full disclosure — in response to trainee skill. Grounded in a structural equation model fit to nearly 2,000 hours of real-world psychotherapy transcripts, the AVP quantifies how therapist empathy and exploration shift patient openness. An LLM generates utterances conditioned on a disclosure level updated each turn by the dynamics module. In evaluation with 20 clinicians and trainees over 80 sessions (1,033 turns), the AVP's disclosure rises in response to therapist empathy and exploration, while a prompt-only baseline stays flat; ablations confirm that the empirically motivated parameterization outperforms alternatives. This represents a significant advance in intelligent-tutoring for professional training domains where authentic practice opportunities are scarce and where existing role-play systems lack grounded, adaptive behavior.
Related Pages
- knowledge-tracing-irt — Knowledge tracing models and IRT for student modeling
- intelligent-tutoring — AI tutoring systems and adaptive instruction
- learning-analytics — Data-driven analysis of learning processes
- ai-literacy — Understanding and evaluating AI tools
- student-modeling — Representing learner knowledge and behavior
- cs-education — Computing education research and pedagogy
- self-regulated-learning — Metacognitive strategies for independent learning
- formative-assessment — Ongoing assessment to inform instruction
- automated-grading — AI-assisted evaluation of student work
- scaffolding — Instructional support that fades with competence
- ai-assessment-human-tutors -- AI-driven assessment of human tutor training performance correlates with real-li...
- astra-atco-training-simulator -- ASTRA uses autonomous AI sim-pilots for scalable air traffic control training; r...
- ai-psychotherapy-training-avatars -- AI-driven interactive patient avatars for psychotherapy training provide accessi...
- medeasy-ai-standardized-patients -- MedEasy multi-agent system simulates standardized patients with varying conditio...
- genai-patient-education-transplant-handbooks — Grounding genAI patient-education assistants in institution-authored handbooks produces inconsistent
Citations
APA: Angela Chen, Siwei Jin, Catherine Bao, Canwen Wang, Robert E. Kraut, Tongshuang Wu, Haiyi Zhu (2026). The Empirically Grounded Adaptive Virtual Patient for Psychotherapy Training. arXiv:2606.10051. cs.CY, cs.HC.