The detection of mental disorders heavily relies on the knowledge and experienceof psychologists. Existing tools used by psychologists and physicians for diagnosing mental disorders have three main challenges: there is no interpretable result provided, for example, brain maps for the visualization purpose; they are time-consuming and need to be administrated by well-trained engineers or by professional clinicians; the interaction between psychologists (or clients) and existing tools demonstrate a lack of efficiency. My visoin is:

Grants

Research Center for Informatics was supported by European Union and completed in Czech Technical University from Sep 2018 to Nov 2024. The goal of RCI is to foster collaboration between the experts in different fields of computer science, between fundamental scientists and application-driven researchers. This project supported my PhD study, and my Duty includes:

• To optimize language models by using neuroscience knowledge ( Tools: Python, SQL);
• To analyze the stability and controllability of Artificial Neural Networks on various optimization components ( Tools: Python, MATLAB);
• To develop a psychotherapy chatbot by using language models ( Tools: Python, JavaScript);

This project aims to understand what information is exchanged, in terms of brain recordings, the nature and the mechanisms of brain computation; to inspire the computer science, in terms of high performance computing, interpretability of AI; to develop new methods of effective modelling of brain dynamics; to gain the knowledge about the dynamics of brain states. This project is supporting my PhD study, and my Duty includes:

• To construct functional brain networks via Granger causality, Transfer Entropy, Directed Information ( Tools: MATLAB, eeglib);
• To bridge the connection between language models and EEG signals via open-source large language models ( Tools: Python, MATLAB);
• To understand hidden specific information transfer dynamics in neural communication ( Tools: Python, MATLAB, C/C++);

Self-motivated

The purpose of this project is to analyze the dynamic of commmon artificial neural netowkrs (ANNs) and then control them. We proposed a new roadmap to analyze the convergence, stability and robustness of ANNs. See my Duty below:

• To conduct the control theory of ANNs;
• To simulate the systematic response of ANNs on various optimizationi components;
• To optimize the halucination of generation models (LLMs, LVMs and vLLMs).

What we found in the first paper and the second paper have proved to benefit the stability of artificial neural networks. This technique provides with:

• a new rule to measure artificial neural networks;
• a practical way to design artificial neural networks and optimization componnents.