Explicitly antisymmetrized neural network layers for variational Monte Carlo simulation

Jeffmin Lin, Gil Goldshlager, Lin Lin

Abstract: Highlights•Explicitly antisymmetrized neural network layers as a diagnostic tool for VMC.•Generic antisymmetric layer with FermiNet backflow can accurately model ground state.•Factorized antisymmetric layer with FermiNet backflow does not outperform FermiNet.•Full 1-determinant FermiNet outperforms 64 standard determinants on dissociating N2.