Kaiwen Tuo

I'm Kaiwen Tuo, an incoming Ph.D. student at the Hong Kong University of Science and Technology (HKUST), advised by Prof. Jiaya Jia. Before joining HKUST, I received my B.E. in Computer Science and Technology from Tongji University (Guohao College). I was honored to be selected as the Undergraduate Valedictorian and deliver a speech at Tongji University’s graduation ceremony.

During my undergraduate studies, my research primarily focused on building multimodal foundation models, spanning evaluation, data curation, and post-training. More specifically, I am interested in: (1) identifying the failure modes of frontier foundation models, (2) constructing high-quality post-training data, and (3) developing effective post-training recipes to push the capability boundaries of multimodal foundation models, particularly in instruction following and agentic RL. I am fortunate to work with Prof. Huan Wang and Prof. Jiaqi Wang.