About

I am a mathematics Ph.D. candidate at Colorado State University expecting to defend in early spring of 2027. I enjoy thinking about how ideas from pure mathematics can advance deep learning.

My research focuses on extending modern equivariant learning architectures to settings in which signals live on Riemannian manifolds. In particular, I’ve extended Weiler’s gauge field theory of convolutional neural nets to general normalized aggregation which covers attention. I have implemented a novel sphere transformer which can process fields of arbitrary type.

See my ICLR 2027 submission under review here.

In parallel, I apply machine learning and statistical methods to extract reliable insight from biological datasets in the small-n, large-p regime.