Circe Hsu

Hello! I am a multidisciplinary researcher working at the intersection of mathematics, physics, and machine learning. Currently, I am a machine learning scientist at Matterworks working on the next generation of foundation models for chemistry applications.

I have previously worked as an undergraduate research assistant in the Geometric Learning Lab at Northeastern University under Robin Walters, where my work combined geometric deep learning methods with problems in physics and math. I have also previously worked as a CIQM research intern in the Kaxiras Group at Harvard University.

Some of my work includes physics-informed neural networks for periodic quantum systems. I am also greatly interested in the intersection of machine learning and mathematics, specifically in developing ML tools for math research. I have also developed equivariant property prediction models to leverage the inherent symmetries of crystals for materials science applications.