Machine learning a fixed point action with a convolutional neural network

by Prof. Urs Wenger (University of Bern)

Wednesday 15 May 2024, 16:00 → 18:00 Europe/Berlin

Fixed point lattice actions based on renormalization group transformations have continuum classical properties unaffected by discretization effects and reduced lattice artifacts at the quantum level. They provide a possible way to extract continuum physics with coarser lattices, thereby allowing to circumvent problems with critical slowing down and topological freezing toward the continuum limit. I describe how we use a gauge-equivariant convolutional neural network and machine learning methods to obtain a fixed point action for four-dimensional SU(3) gauge theory. The large operator space allows us to find superior parametrizations compared to previous studies, a necessary first step for future Monte Carlo simulations.