Published multiple papers in top-tier conferences such as ICCV, NeurIPS, ICML, ACL, etc.; Organized several workshops at international conferences; Served as an Area Chair for ICLR 2026; Released open-source projects like MobileLLM-R1, MiniGPT-v2, etc.
Research Experience
Visiting scholar at Carnegie Mellon University (Sept 2019 - Dec 2021), advised by Prof. Marios Savvides and Prof. Eric Xing; Joined Meta Inc. as a Research Scientist (Dec 2021).
Education
Ph.D. from Hong Kong University of Science and Technology (June 2021), supervised by Prof. Kwang-Ting Tim CHENG; Bachelor's degree from Fudan University (June 2016).
Background
Currently a Staff Research Scientist and Tech Lead at Meta. Research focuses on improving the efficiency and deployability of foundation models through architectural optimization, low-bit quantization, sparsity, and more. Specifically, interested in using deep learning to solve practical problems in the industry.