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

TLDR

We have content for each. Let's aggregate: Purpose, Mechanism: "Abstract 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." Background: "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." Findings: multiple sentences: state-of-the-art accuracy, data efficiency, outperforming with up to three orders fewer training data, challenging belief, high data efficiency allows accurate potentials using high-order quantum chemical reference and enables high-fidelity MD over long times. Combine: "NequIP achieves state‑of‑the‑art accuracy on diverse molecules and materials, outperforms existing models with up to three orders of magnitude fewer training data, and enables accurate, long‑timescale molecular‑dynamics simulations using high‑order quantum‑chemical references." Make sure it's one sentence. Yes.

Abstract

Abstract 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.