Browse publications on Google Scholar (top-right) ↗
Resume (English only)
Academic Achievements
Preprints: 'Semantic-Clipping: Efficient Vision-Language Modeling with Semantic-Guided Visual Selection' etc.; Publications: 'Code Execution as Grounded Supervision for LLM Reasoning' (EMNLP 2025), 'T-REG: Preference Optimization with Token-Level Reward Regularization' (ACL 2025) etc.
Research Experience
Research scientist at Google Deepmind.
Education
Ph.D. from the University of Southern California, under the mentorship of Prof. Muhao Chen; Bachelor's degree in Computer Science and Applied Mathematics from HKUST, graduated in 2014.
Background
Research interests: natural language processing and machine learning, with a particular emphasis on the post-training of large language models. Interested in instruction tuning, preference modeling, and reinforcement learning from human feedback.