I am an Assistant Professor in the Department of CIS at the University of Pennsylvania. I also work part-time as a Staff Research Scientist at Apple (MLR) and was a full-time researcher since 2022. Prior to joining Apple, I was a Research Scientist at Meta AI (FAIR Labs).
I obtained my Ph.D. degree at the department of Electrical and Electronic Engineering, University of Hong Kong in 2018 and I was supervised by Prof. Victor O.K. Li. My doctoral work explored efficient neural machine translation systems. I also spent a wonderful time visiting the CILVR Lab, New York University working with Prof. Kyunghyun Cho. Before that, I obtained my Bachelor’s degree at the Electronic Engineering Department, Tsinghua University in 2014.
I will be recruiting self-motivated PhD students, visiting students and research interns for this graduate application cycle. I encourage to use this form.
My research goal is to advance the capabilities of AI agents to interact with the physical world, with a special emphasis on leveraging generative machine learning approaches for world modeling, iterative reasoning, and effective decision-making. I am particularly interested in developing multi-modal generative models that span natural language, images, videos, and 3D, ensuring these models are Efficient, Flexible, Scalable, and Knowledgeable. By bridging foundational machine learning research with real-world applications, my research seeks to create robust AI systems capable of operating seamlessly in complex and dynamic environments.
One paper was accepted by ICLR 2025..
5 Jan, 2024One paper was accepted by TMLR..
11 Nov, 2024I will be joining UPenn as an assistant professor in CIS department starting Fall 2025!
26 Sep, 2024One paper was accepted by NeurIPS 2024..
The Thirty-Eighth Annual Conference on Neural Information Processing Systems (NeurIPS)
Vancouver, Canada
CONFERENCE ON LANGUAGE MODELING (COLM)
Philadelphia, USA
The Forty-First International Conference on Machine Learning (ICML)
Vienna, Austria
The Twelve International Conference on Learning Representations (ICLR)
Vienna, Austria