Nate MacFadden

My research:

I work in the Liam McAllister group. We study compactifications of string theory on Calabi-Yau threefolds using the software CYTools. This software boils down to efficient algorithms on regular triangulations of reflexive polytopes and optimization in high-dimensional polyhedral fans. I develop/maintain this software.

My expertise is:

Regular triangulations of reflexive polytopes. Convex cones. Polyhedral fans.

A problem I’m grappling with:

Finding (as many as possible) roots of highly structured functions in O(100) dimensions, as quickly as possible.

I’ve got my eyes on:

Wrapping up existing projects so I can focus on said root finding.

I want to know more about:

Learning ML, specifically in areas where it outperforms non-ML algorithms. Specific areas of interest are reinforcement learning and transformers.