TeaNet: universal neural network interatomic potential inspired by iterative electronic relaxations

Abstract

A universal interatomic potential for an arbitrary set of chemical elements is urgently needed in computational materials science. Graph convolution neural network (GCN) has rich expressive power, but previously was mainly employed to transport scalars and vectors, not rank $\ge 2$ tensors. As classic interatomic potentials were inspired by tight-binding electronic relaxation framework, we want to represent this iterative propagation of rank $\ge 2$ tensor information by GCN. Here we propose an architecture called the tensor embedded atom network (TeaNet) where angular interaction is translated into graph convolution through the incorporation of Euclidean tensors, vectors and scalars. By applying the residual network (ResNet) architecture and training with recurrent GCN weights initialization, a much deeper (16 layers) GCN was constructed, whose flow is similar to an iterative electronic relaxation. Our traning dataset is generated by density functional theory calculation of mostly chemically and structurally randomized configurations. We demonstrate that arbitrary structures and reactions involving the first 18 elements on the periodic table (H to Ar) can be realized satisfactorily by TeaNet, including C-H molecular structures, metals, amorphous SiO${}_2$, and water, showing surprisingly good performance (energy mean absolute error 19 meV/atom) and robustness for arbitrary chemistries involving elements from H to Ar.

Revision: Add detailed network description

References in corpus (8)

• Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translation
• WaveNet: A Generative Model for Raw Audio
• E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate Interatomic Potentials
• SchNet: A continuous-filter convolutional neural network for modeling quantum interactions
• SpookyNet: Learning Force Fields with Electronic Degrees of Freedom and Nonlocal Effects
• FCHL revisited: faster and more accurate quantum machine learning
• Equivariant message passing for the prediction of tensorial properties and molecular spectra
• Semi-supervised Knowledge Transfer for Deep Learning from Private Training Data

Cited by in corpus (21)

• Towards Universal Neural Network Potential for Material Discovery Applicable to Arbitrary Combination of 45 Elements
• General-purpose machine-learned potential for 16 elemental metals and their alloys
• A practical guide to machine learning interatomic potentials -- Status and future
• Evaluation of the MACE Force Field Architecture: from Medicinal Chemistry to Materials Science
• Machine learning in nuclear materials research
• The Evolution of Machine Learning Potentials for Molecules, Reactions and Materials
• Force-Field-Enhanced Neural Network Interactions: from Local Equivariant Embedding to Atom-in-Molecule properties and long-range effects
• Tensor-reduced atomic density representations
• Neural Network Kinetics for Exploring Diffusion Multiplicity and Chemical Ordering in Compositionally Complex Materials
• E(n)-Equivariant Cartesian Tensor Passing Potential
• Decoding the hidden dynamics of super-Arrhenius hydrogen diffusion in multi-principal element alloys via machine learning
• FeNNol: an Efficient and Flexible Library for Building Force-field-enhanced Neural Network Potentials
• Facilitating {\it ab initio} configurational sampling of multicomponent solids using an on-lattice neural network model and active learning
• Fitting to magnetic forces improves the reliability of magnetic Moment Tensor Potentials
• Automated atomistic simulations of dissociated dislocations with ab initio accuracy
• Multi-head committees enable direct uncertainty prediction for atomistic foundation models
• Equivariant Tensor Network Potentials
• Weisfeiler-Lehman Embedding for Molecular Graph Neural Networks
• Exact average many-body interatomic interaction model for random alloys
• Symmetry-adapted graph neural networks for constructing molecular dynamics force fields
• Revealing trends in catalytic activity of adatoms for hydrogen adsorption on carbon: a case study of graphene and carbon nanotube