Neural Network Potentials: A Concise Overview of Methods

TL;DR

This paper reviews neural network-based machine learning potentials, highlighting their methods, differences, and challenges in modeling atomic systems for large-scale simulations across chemistry, physics, and materials science.

Contribution

It provides a concise overview of neural network potentials, emphasizing conceptual differences and open challenges in the field.

Findings

Neural network potentials enable large-scale atomistic simulations.

Different descriptors and long-range interactions are key considerations.

Open challenges include system dimensionality and non-local phenomena.

Abstract

In the past two decades, machine learning potentials (MLP) have reached a level of maturity that now enables applications to large-scale atomistic simulations of a wide range of systems in chemistry, physics and materials science. Different machine learning algorithms have been used with great success in the construction of these MLPs. In this review, we discuss an important group of MLPs relying on artificial neural networks to establish a mapping from the atomic structure to the potential energy. In spite of this common feature, there are important conceptual differences, which concern the dimensionality of the systems, the inclusion of long-range electrostatic interactions, global phenomena like non-local charge transfer, and the type of descriptor used to represent the atomic structure, which can either be predefined or learnable. A concise overview is given along with a discussion of the open challenges in the field.