I’m a second-year PhD @CityUHK, advised by Ning Miao. Previously, I worked as a master student @Fudan DISC (affiliated to FudanNLP), advised by Zhongyu Wei and Siyuan Wang. I was a research intern @XLANG, advised by Tao Yu. I earned my bachelor’s degree from @Beijing Institute of Technology.
My research focuses on challenging problem design (VHG, DS-1000, ARKS), rewards (ALaRM, HAF-RM), and evaluation (DS-1000, SSAE). I’m particularly interested in how challenging problems and reliable evaluation signals can support scalable oversight as models surpass human experts.