CCD-CBT: Multi-Agent Therapeutic Interaction for CBT Guided by Cognitive Conceptualization Diagram
This work addresses the need for scalable mental-health support tools by providing a more clinically-plausible simulation of therapy interactions, though it is incremental in advancing existing CBT simulation methods.
The paper tackles the problem of simulating Cognitive Behavioral Therapy (CBT) counselors using large language models by addressing limitations in static profiles and omniscient interactions, resulting in a multi-agent framework that improves counseling fidelity and positive-affect enhancement, as shown by evaluations with clinical scales and expert therapists.
Large language models show potential for scalable mental-health support by simulating Cognitive Behavioral Therapy (CBT) counselors. However, existing methods often rely on static cognitive profiles and omniscient single-agent simulation, failing to capture the dynamic, information-asymmetric nature of real therapy. We introduce CCD-CBT, a multi-agent framework that shifts CBT simulation along two axes: 1) from a static to a dynamically reconstructed Cognitive Conceptualization Diagram (CCD), updated by a dedicated Control Agent, and 2) from omniscient to information-asymmetric interaction, where the Therapist Agent must reason from inferred client states. We release CCDCHAT, a synthetic multi-turn CBT dataset generated under this framework. Evaluations with clinical scales and expert therapists show that models fine-tuned on CCDCHAT outperform strong baselines in both counseling fidelity and positive-affect enhancement, with ablations confirming the necessity of dynamic CCD guidance and asymmetric agent design. Our work offers a new paradigm for building theory-grounded, clinically-plausible conversational agents.