Machine learning for lattice field theory

by Gurtej Kanwar (Uni Bern)

Tuesday 22 Nov 2022, 16:00 → 18:00 Europe/Zurich

Zoom (CERN)

State-of-the-art lattice calculations are limited by the enormous computational effort required for precision studies, motivating developments to lattice methods that reduce such costs. I will discuss recent progress in applying machine learning to lattice field theory in order to drastically improve efficiency. A key feature of this approach is using machine learning components in frameworks where exactness of the physics results is not compromised. Incorporating gauge symmetry exactly in the machine learning architectures is also demonstrated to significantly improve the efficacy of the method in gauge theories.