E(3)-equivariant graph neural networks for data-efficient and accurate interatomic potentials

An E(3)-equivariant deep learning interatomic potential is introduced for accelerating molecular dynamics simulations. The method obtains state-of-the-art accuracy, can faithfully describe dynamics of complex systems with remarkable sample efficiency. This work presents Neural Equivariant Interatomic Potentials (NequIP), an E(3)-equivariant neural network approach for learning interatomic potentials from ab-initio calculations for molecular dynamics simulations. While most contemporary symmetry-aware models use invariant convolutions and only act on scalars, NequIP employs E(3)-equivariant convolutions for interactions of geometric tensors, resulting in a more information-rich and faithful representation of atomic environments. The method achieves state-of-the-art accuracy on a challenging and diverse set of molecules and materials while exhibiting remarkable data efficiency. NequIP outperforms existing models with up to three orders of magnitude fewer training data, challenging the widely held belief that deep neural networks require massive training sets. The high data efficiency of the method allows for the construction of accurate potentials using high-order quantum chemical level of theory as reference and enables high-fidelity molecular dynamics simulations over long time scales.

Automated summary

This paper introduces an E(3)-equivariant deep learning method for accelerating molecular dynamics simulations. The method shows state-of-the-art accuracy and remarkable sample efficiency in faithfully describing the dynamics of complex systems. The Neural Equivariant Interatomic Potentials (NequIP) approach employs E(3)-equivariant convolutions to interact with geometric tensors, resulting in a more information-rich and faithful representation of atomic environments. NequIP outperforms existing models with significantly fewer training data, challenging the commonly held belief about the necessity of massive training sets for deep neural networks.