Chenxu Liu 刘宸绪

My Ph.D. research focuses on quality assurance and testing of graphical user interfaces (GUI), especially web GUI testing. I integrate traditional algorithms with deep learning models, large language models, and GUI agents to improve testing effectiveness.

My current and future work focuses on Code Agent and GUI Agent, incorporating knowledge of software engineering and software testing to evaluate and build practical, reliable agents for real-world use. My research includes open-ended evaluation of the front-end coding capabilities of models (WebDev and Office), as well as evaluation of User Agents and GUI Agents in realistic usage scenarios.