About

Linrong Cai is an M.S.E. master’s student in Computer Science at Princeton University, advised by Professor Zhuang Liu. His research centers on open reinforcement-learning recipes for vision-language models. He is one of two project leads on Vero, a fully open RL recipe that improves vision-language reasoning across diverse tasks; the resulting paper was selected as an Oral at the CVPR 2026 DataMFM workshop and has been submitted to ECCV 2026. He also mentored three Princeton undergraduates on OccludeBench, a benchmark on visual grounding in VLMs accepted at the CVPR 2026 ViSCALE workshop. Prior to Princeton, Linrong received his B.S. in Computer Science with Honors and his B.S. in Mathematics from the University of Wisconsin–Madison, where he was an equal-contribution first author at NeurIPS 2024 and a co-author at ICLR 2024 with Professor Frederic Sala. He also interned as a Quantitative Research Intern at Orient Securities.

Education

Princeton University

Ph.D. , Computer Science

Class of 2027

Awards & Publications