CS PhD student at Duke University, co-advised by Prof. Shuyan Zhou and Prof. Bhuwan Dhingra. Previously at Peking University (Prof. Xiaojun Wan), Microsoft Research, and UCSB NLP Group (Prof. William Yang Wang). I study how AI systems can recursively self-improve over long horizons to solve open-ended problems, with a focus on self-referential agents and world models. Feel free to reach out to me if you'd like to discuss research or explore potential collaboration!