My research focuses on Agentic RL and foundation models, with a particular emphasis on developing trustworthy intelligent agents that can effectively interact with real-world environments.

I am the creator of WizardCoder and have worked on a range of widely recognized projects across language, multimodal, and agentic AI, including VideoLLaMA2, ScreenSpot-Pro, MCP-Universe, and StateAct (SFR-CUA). I was also deeply involved in the development of the open-source WizardLM project and contributed to Mercury, the first commercial-scale diffusion language model. My work has resulted in more than 30 publications at leading AI conferences, including ICLR, CVPR, ICCV, ACL, and EMNLP.

I obtained my PhD in Computer Science from Hong Kong Baptist University (2025). Previously, I was a visiting researcher at the National University of Singapore (2024), collaborating with Prof. Mohan Kankanhalli . I have worked at Salesforce Research, Microsoft Research and Alibaba. For research collaboration or other inquiries, please visit the contact page or reach out via WeChat (ID: chiyeunglaw).