Speaker
Description
Classically perfect fixed-point (FP) actions based on the renormalization group allow to reliably extract continuum physics from Monte Carlo simulations at coarse lattice spacings, thereby avoiding topological freezing. While these FP actions are very complicated, machine-learned gauge-equivariant neural networks enable accurate parametrizations and efficient simulations. In this talk I present our latest results from simulations of such a machine-learned FP action for 4-dimensional SU(3) gauge theory. In particular I discuss the continuum limits of some thermodynamical properties of the deconfinement phase transition, such as the latent heat and the interface tension, and the topological susceptibility based on a machine-learned FP topological charge operator.