Weichi earned her Ph.D. in Statistics from New York University, where her dissertation focused on adapting machine-learning models to address challenging problems in medical and physical sciences. As a Schmidt AI in Science Fellow, she was developing generative models for molecular design under data-scarce conditions to advance scientific discovery in chemistry and materials science. Her research leveraged cutting-edge architectures—including flow matching, diffusion models, and transformer-based frameworks originally designed for image and text processing—to generate novel chemical structures with desired properties. Recognizing that these generative approaches often required abundant, representative datasets and might have struggled when target properties fell outside the observed training distribution, Weichi tackled the challenge of limited data through model architecture refinement, statistical methodology innovation, and strategic experimental design.
- AI Mentor: Yixin Wang, Statistics, LSA
- Science Mentor: Bryan Goldsmith, Chemical Engineering, College of Engineering