I am a PhD student in Electrical and Computer Engineering at the University of British Columbia. My research combines generative modeling and reinforcement learning with medical imaging, mathematical optimization, multi agent systems, and computational design.
- RRT (paper) using item response theory to reduce the judge cost during rubric based LLM training.
- MOTFM (paper) develops optimal transport flow matching for fast medical image synthesis in 2D and 3D.
- EvoCut (paper) combines language models with evolutionary search to generate acceleration cuts for mixed integer programs.
- MASPRM (paper) is a process reward model that guides search in multi agent systems.
- RL-Kirigami (journal · arXiv) combines conditional flow matching and reinforcement learning for inverse kirigami design and rapid fabrication.
- Incorporating vision-language synthetic data enhances performance on echocardiography downstream tasks Journal of Medical Imaging (2026).