# DeepSAGE: Stage-Aware Reinforcement Learning for Structured CBT Counseling Dialogue

arXiv: https://arxiv.org/abs/2608.22615 · HTML: https://mlsanity.com/paper/2608.22615/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Clinical and biomedical applications, one of 17 directions in LLM reasoning.
- Builds on 10 of its 45 references that are in the map.
- First posted on arXiv 23 Aug 2026.

## Placed in

- LLM reasoning / [Clinical and biomedical applications](https://mlsanity.com/lr/clinical-and-biomedical-applications/)

## Builds on (most cited first, 10 of 10)

- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [Deep Reinforcement Learning for Dialogue Generation](https://mlsanity.com/paper/1606.01541/) (2016)
- [The Design and Implementation of XiaoIce, an Empathetic Social Chatbot](https://mlsanity.com/paper/1812.08989/) (2018)
- [CBT-LLM: A Chinese Large Language Model for Cognitive Behavioral Therapy-based Mental Health Question Answering](https://mlsanity.com/paper/2403.16008/) (2024)
- [If I Hear You Correctly: Building and Evaluating Interview Chatbots with Active Listening Skills](https://mlsanity.com/paper/2002.01862/) (2020)
- [AutoCBT: An Autonomous Multi-agent Framework for Cognitive Behavioral Therapy in Psychological Counseling](https://mlsanity.com/paper/2501.09426/) (2025)
- [Strategic Dialogue Management via Deep Reinforcement Learning](https://mlsanity.com/paper/1511.08099/) (2015)
- [GoChat: Goal-oriented Chatbots with Hierarchical Reinforcement Learning](https://mlsanity.com/paper/2005.11729/) (2020)
- [Neural Generation Meets Real People: Building a Social, Informative Open-Domain Dialogue Agent](https://mlsanity.com/paper/2207.12021/) (2022)
- [NLP meets psychotherapy: Using predicted client emotions and self-reported client emotions to measure emotional coherence](https://mlsanity.com/paper/2211.12512/) (2022)

## Abstract (arXiv)

Large Language Model (LLM)-based counseling agents can generate fluent and supportive responses, but they often lack the structured, goal-directed progression required to conduct a coherent therapeutic session. We present DeepSAGE (Strategic AI Guidance Engine), a hybrid LLM--Deep Reinforcement Learning (DRL) framework for stage-aware counseling dialogue grounded in the first session of Cognitive Behavioral Therapy (CBT). DeepSAGE represents the session as eleven stages with explicit therapeutic objectives, with an external controller determines stage completion and the DRL model selects therapeutic intentions that guide LLM response generation. We evaluate DeepSAGE against six retrieval-, prompting-, stage-, and policy-based alternatives. DeepSAGE elicits higher simulated client engagement and openness and achieves the strongest balance of stage-goal completion and dialogue efficiency among stage-structured systems. Domain expert review further indicates that the generated conversations exhibit broadly plausible emotional trajectories and recognizable CBT processes. Because the evaluation relies primarily on simulated clients and model-based metrics, these findings demonstrate comparative dialogue-control improvements rather than clinical effectiveness. These results suggest that combining stage-structured dialogue with learned strategy selection is a promising approach for AI counseling, though clinical effectiveness, safety, and real-world utility require further human evaluation.