We develop machine learning methods for understanding molecular structure and function. Our research focuses on graph neural networks, generative modeling, representation learning, and property prediction across chemical spaces.
A major goal is designing unified models capable of transferring chemical intuition across diverse molecular tasks.