Search papers, labs, and topics across Lattice.
The paper introduces STEP, a dataset modeling Cognitive Behavioral Therapy (CBT) counseling dialogues with explicit automatic thoughts and action-level counseling sequences. Using STEP, they train STEPPER, a counseling agent that proactively elicits automatic thoughts and executes CBT interventions. They then refine STEPPER using preference learning from simulated counseling sessions, demonstrating improved clinical grounding, coherence, and personalization compared to baselines.
Forget generic chatbot empathy – this agent proactively elicits and addresses negative thoughts with clinically-grounded CBT interventions.
Cognitive Behavioral Therapy (CBT) aims to identify and restructure automatic negative thoughts pertaining to involuntary interpretations of events, yet existing counseling agents struggle to identify and address them in dialogue settings. To bridge this gap, we introduce STEP, a dataset that models CBT counseling by explicitly reflecting automatic thoughts alongside dynamic, action-level counseling sequences. Using this dataset, we train STEPPER, a counseling agent that proactively elicits automatic thoughts and executes cognitively grounded interventions. To further enhance both decision accuracy and empathic responsiveness, we refine STEPPER through preference learning based on simulated, synthesized counseling sessions. Extensive CBT-aligned evaluations show that STEPPER delivers more clinically grounded, coherent, and personalized counseling compared to other strong baseline models, and achieves higher counselor competence without inducing emotional disruption.
