Compact Kernel/Neural Network Representation for Accurate, Fast, and Global Reactive Molecular Potential Energy Surfaces

Abstract

Atomistic simulations are a powerful tool for studying the dynamics of molecules, proteins, and materials on wide time and length scales. Their reliability and predictive power depend directly on the accuracy of the underlying potential energy surface (PES). This work introduces KerNN, a combined kernel- and neural network-based approach to represent molecular PESs. Compared to state-of-the-art neural network PESs, the number of learnable parameters of KerNN is significantly reduced. This speeds up training and evaluation times by several orders of magnitude, while retaining high prediction accuracy. Importantly, using kernels as the features also improves the extrapolation capabilities of KerNN far beyond the coverage provided by the training data, which solves a fundamental problem of neural-network-based PESs. KerNN applied to spectroscopy and reaction dynamics shows excellent performance on observables, including vibrational bands computed from classical and quantum simulations.