Research

During my PhD, I have focused on understanding the theoretical limits of equivariant neural networks. My work includes studying how to handle different forms of symmetry breaking, analyzing asymptotic trade-offs between runtime and expressivity for core operations, and investigating loss landscape geometry of equivariant networks. Recently, I have also become interested in neural quantum states and the representation of antisymmetric functions.

More broadly, I am driven by a fundamental curiosity. I enjoy the process of distilling the essence of some phenomenon in a complex system, and then thoroughly analyzing, understanding, and generalizing it.

Publications

Please also see my Google Scholar.