Solving the Schr"odinger Equation with a Graph Neural Network

Ab-Initio Potential Energy Surfaces by Pairing GNNs with Neural Wave Functions

We combine a graph neural network (gnn) with a neuralwave function to simultaneously solve the schr\"odinger equation for multiple molecular geometries via variational monte-carlo (vmc) framework.This enables us to model continuous subsets of the potential energy surface with a single training pass.Compared to existing state-of-the-art networks, our potential energy surface network (pesnet) speedsup training for multiple geometries by up to 40 times while matching orsurpassing their accuracy.This may open the path to accurate and orders of magnitude cheaper quantum mechanical calculations.