What is PsycoLLM
PsycoLLM is a Chinese psychological large language model launched by the School of Computer Science and Information Engineering at Hefei University of Technology. It is trained on a high-quality psychological dataset to enhance the understanding and assessment of mental health issues. The model's dataset covers single-turn question answering, multi-turn dialogues, and knowledge-based question answering, using an innovative data generation and optimization process to ensure the authenticity and applicability of the data. PsycoLLM performs excellently in multi-dimensional psychological benchmark tests such as professional ethics, theoretical knowledge, and case analysis. Compared with other models, it demonstrates stronger performance and more accurate judgment capabilities, providing strong technical support for research and applications in the field of mental health.
Main Functions of PsycoLLM
- Understanding and Answering Psychological Questions: Accurately understand the psychological questions raised by users and provide professional and accurate answers to help users obtain psychological support and guidance.
- Multi-turn Dialogue Interaction: Support multi-turn dialogues with users, and based on continuous question-answer interactions, gain an in-depth understanding of users' psychological states and needs, providing more targeted suggestions and help.
- Popularization and Education of Psychological Knowledge: Use a rich knowledge base of psychology to popularize mental health knowledge among users and improve users' awareness of psychological issues and self-regulation ability.
- Emotion Recognition and Support: Recognize users' emotional states, such as anxiety and depression, and provide corresponding emotional support and comfort to help users relieve emotional distress.
- Mental Health Assessment and Suggestions: Conduct a preliminary assessment of users' mental health status and provide corresponding suggestions based on the assessment results, such as seeking professional psychological counseling and engaging in self-regulation.
Technical Principles of PsycoLLM
- Training on High-Quality Datasets: PsycoLLM is trained on a high-quality psychological dataset. The dataset includes various types of data such as single-turn question answering, multi-turn dialogues, and knowledge-based question answering, covering rich psychological knowledge and real psychological counseling scenarios, enabling the model to learn professional psychological knowledge and dialogue techniques.
- Multi-step Data Generation and Optimization Process: In the process of generating multi-turn dialogue data, a multi-step process is used, including multi-turn question-answer generation, evidence judgment, and dialogue optimization. First, preliminary multi-turn dialogues are generated, then it is determined whether each answer in the dialogue is supported by evidence, and finally the dialogue is optimized to improve its coherence, authenticity, and applicability.
- Supervised Fine-Tuning: On the basis of the pre-trained model, supervised fine-tuning is used to further improve the model's performance in the field of psychology. During the fine-tuning process, the model is trained on a high-quality psychological dataset to better understand and generate psychology-related text.
- Transformer Architecture: The Transformer architecture is used as the core model structure, using the self-attention mechanism to capture long-distance dependencies in text and achieve efficient understanding and generation of text.
Project Address of PsycoLLM
- GitHub Repository：https://github.com/MACLAB-HFUT/PsycoLLM
- HuggingFace Model Library：https://huggingface.co/MACLAB-HFUT/PsycoLLM
- arXiv Technical Paper：https://arxiv.org/pdf/2407.05721
Application Scenarios of PsycoLLM
- Personal Mental Health Support: When users are in a low mood, they can converse with PsycoLLM to obtain emotional support and adjustment suggestions, helping relieve stress and restore emotional balance.
- Psychological Counseling: Users describe psychological issues before psychological counseling, generating a pre-assessment report to provide reference information for counselors and improve counseling efficiency.
- Student Mental Health Education: PsycoLLM assists teaching in mental health courses, explaining psychological knowledge, helping students understand and master skills such as emotion management, and improving mental health literacy.
- Community Mental Health Services: Community residents receive psychological support and counseling services, resolve psychological issues in life, and promote community harmony.