I work primarily on AI safety. My main research interests include AI alignment, human-GenAI interaction, and mechanistic interpretability. My goal is to understand how humans can collaborate effectively with AI systems, which requires first understanding how those systems actually work.