Psyche-R1: Towards Reliable Psychological LLMs through Unified Empathy, Expertise, and Reasoning

TL;DR

Psyche-R1 is a Chinese psychological large language model that combines empathy, expertise, and reasoning through a novel data pipeline and hybrid training, achieving reliable responses in mental health applications.

Contribution

This paper introduces Psyche-R1, the first Chinese psychological LLM integrating empathy, expertise, and reasoning with a new data synthesis and training strategy.

Findings

Achieves comparable performance to much larger models on psychological benchmarks.

Utilizes a novel data pipeline generating 75k question-rationale pairs.

Employs a hybrid training strategy improving reasoning and empathetic responses.

Abstract

Amidst a shortage of qualified mental health professionals, the integration of large language models (LLMs) into psychological applications offers a promising way to alleviate the growing burden of mental health disorders. Recent reasoning-augmented LLMs have achieved remarkable performance in mathematics and programming, while research in the psychological domain has predominantly emphasized emotional support and empathetic dialogue, with limited attention to reasoning mechanisms that are beneficial to generating reliable responses. Therefore, in this paper, we propose Psyche-R1, the first Chinese psychological LLM that jointly integrates empathy, psychological expertise, and reasoning, built upon a novel data curation pipeline. Specifically, we design a comprehensive data synthesis pipeline that produces over 75k high-quality psychological questions paired with detailed rationales, generated through chain-of-thought (CoT) reasoning and iterative prompt-rationale optimization, along with 73k empathetic dialogues. Subsequently, we employ a hybrid training strategy wherein challenging samples are identified through a multi-LLM cross-selection strategy for group relative policy optimization (GRPO) to improve reasoning ability, while the remaining data is used for supervised fine-tuning (SFT) to enhance empathetic response generation and psychological domain knowledge. Extensive experiment results demonstrate the effectiveness of the Psyche-R1 across several psychological benchmarks, where our 7B Psyche-R1 achieves comparable results to 671B DeepSeek-R1.