MACE: Higher order equivariant message passing neural networks for fast and accurate force fields

I Batatia, DP Kovacs, G Simm… - Advances in Neural …, 2022 - proceedings.neurips.cc
Creating fast and accurate force fields is a long-standing challenge in computational
chemistry and materials science. Recently, Equivariant Message Passing Neural Networks …

Learning local equivariant representations for large-scale atomistic dynamics

A Musaelian, S Batzner, A Johansson, L Sun… - Nature …, 2023 - nature.com
A simultaneously accurate and computationally efficient parametrization of the potential
energy surface of molecules and materials is a long-standing goal in the natural sciences …

Construction of high accuracy machine learning interatomic potential for surface/interface of nanomaterials—A review

K Wan, J He, X Shi - Advanced Materials, 2024 - Wiley Online Library
The inherent discontinuity and unique dimensional attributes of nanomaterial surfaces and
interfaces bestow them with various exceptional properties. These properties, however, also …

Data generation for machine learning interatomic potentials and beyond

M Kulichenko, B Nebgen, N Lubbers, JS Smith… - Chemical …, 2024 - ACS Publications
The field of data-driven chemistry is undergoing an evolution, driven by innovations in
machine learning models for predicting molecular properties and behavior. Recent strides in …

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

S Batzner, A Musaelian, L Sun, M Geiger… - Nature …, 2022 - nature.com
Abstract This work presents Neural Equivariant Interatomic Potentials (NequIP), an E (3)-
equivariant neural network approach for learning interatomic potentials from ab-initio …

DeePMD-kit v2: A software package for deep potential models

J Zeng, D Zhang, D Lu, P Mo, Z Li, Y Chen… - The Journal of …, 2023 - pubs.aip.org
DeePMD-kit is a powerful open-source software package that facilitates molecular dynamics
simulations using machine learning potentials known as Deep Potential (DP) models. This …

Forces are not enough: Benchmark and critical evaluation for machine learning force fields with molecular simulations

X Fu, Z Wu, W Wang, T **e, S Keten… - arxiv preprint arxiv …, 2022 - arxiv.org
Molecular dynamics (MD) simulation techniques are widely used for various natural science
applications. Increasingly, machine learning (ML) force field (FF) models begin to replace ab …

GPUMD: A package for constructing accurate machine-learned potentials and performing highly efficient atomistic simulations

Z Fan, Y Wang, P Ying, K Song, J Wang… - The Journal of …, 2022 - pubs.aip.org
We present our latest advancements of machine-learned potentials (MLPs) based on the
neuroevolution potential (NEP) framework introduced in Fan et al.[Phys. Rev. B 104, 104309 …

The design space of E (3)-equivariant atom-centred interatomic potentials

I Batatia, S Batzner, DP Kovács, A Musaelian… - Nature Machine …, 2025 - nature.com
Molecular dynamics simulation is an important tool in computational materials science and
chemistry, and in the past decade it has been revolutionized by machine learning. This rapid …

[HTML][HTML] Evaluation of the MACE force field architecture: From medicinal chemistry to materials science

DP Kovács, I Batatia, ES Arany… - The Journal of Chemical …, 2023 - pubs.aip.org
The MACE architecture represents the state of the art in the field of machine learning force
fields for a variety of in-domain, extrapolation, and low-data regime tasks. In this paper, we …