LeapFrogLayers: A Trainable Framework for Effective Topological Sampling

29 Jul 2021, 14:45
15m
Oral presentation Algorithms (including Machine Learning, Quantum Computing, Tensor Networks) Algorithms (including Machine Learning, Quantum Computing, Tensor Networks)

Speaker

Sam Foreman (Argonne National Laboratory)

Description

We introduce LeapfrogLayers, an invertible neural network architecture that can be trained to efficiently sample the topology of a 2D U(1) lattice gauge theory. We show an improvement in the integrated autocorrelation time of the topological charge when compared with traditional HMC, and propose methods for scaling our model to larger lattice volumes.

Primary authors

Sam Foreman (Argonne National Laboratory) Xiao-Yong Jin (Argonne National Laboratory) James C Osborn (Argonne ALCF)

Presentation materials