Hello! I'm Shei Chua (Vincent) from Malaysia! I am a Master's candidate in TSAIL lab at Tsinghua University, supervised by Xiaolin Hu.

My research interests lie broadly in AI safety. My earlier work revolves around building and breaking AI safeguards, red-teaming, safety alignment, and model interpretability. More recently, I am interested in reliable AI healthcare systems and real-world safety threats of AI agents, including prompt injection and adversarial vulnerabilities in deployed systems.