Pith. sign in

REVIEW 21 cited by

MACE: Higher Order Equivariant Message Passing Neural Networks for Fast and Accurate Force Fields

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2206.07697 v2 pith:OIIXKO3J submitted 2022-06-15 stat.ML cond-mat.mtrl-scics.LGphysics.chem-ph

Ilyes Batatia , Dávid Péter Kovács , Gregor N. C. Simm , Christoph Ortner , Gábor Csányi This is my paper
classification stat.MLcond-mat.mtrl-scics.LGphysics.chem-ph
keywords messagesequivariantfasthighermessagempnnsorderpassing
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Creating fast and accurate force fields is a long-standing challenge in computational chemistry and materials science. Recently, several equivariant message passing neural networks (MPNNs) have been shown to outperform models built using other approaches in terms of accuracy. However, most MPNNs suffer from high computational cost and poor scalability. We propose that these limitations arise because MPNNs only pass two-body messages leading to a direct relationship between the number of layers and the expressivity of the network. In this work, we introduce MACE, a new equivariant MPNN model that uses higher body order messages. In particular, we show that using four-body messages reduces the required number of message passing iterations to just two, resulting in a fast and highly parallelizable model, reaching or exceeding state-of-the-art accuracy on the rMD17, 3BPA, and AcAc benchmark tasks. We also demonstrate that using higher order messages leads to an improved steepness of the learning curves.

Discussion (0). Sign in to comment.

Forward citations

Cited by 21 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. OpenAlex reports about 287 citations worldwide. Full citation record

  1. A fast summation method for the DFT-D3 dispersion correction

    physics.comp-ph 2026-07 conditional novelty 7.0 of 10

    FourierD3 restores the separability of environment-dependent D3 dispersion coefficients via a functional low-rank decomposition, enabling particle-mesh evaluation of the dispersion sum without a real-space cutoff.

  2. DeepHartree: A Poisson-Coupled Neural Field for One-Shot Density Functional Theory

    physics.chem-ph 2026-04 unverdicted novelty 7.0 of 10

    DeepHartree is a Poisson-coupled E(3)-equivariant neural field that predicts consistent real-space densities and Hartree potentials to accelerate LCAO DFT with near-linear scaling and zero-shot transfer to larger systems.

  3. MOFSimBench: Evaluating Universal Machine Learning Interatomic Potentials In Metal--Organic Framework Molecular Modeling

    cond-mat.mtrl-sci 2025-07 conditional novelty 7.0 of 10

    A new open benchmark shows that top universal machine learning interatomic potentials outperform classical force fields and a fine-tuned MOF-specific potential for most nanoporous materials modeling tasks.

  4. Rem3Di: Learning smooth, chiral 3D molecular descriptors from atomistic foundation models

    physics.chem-ph 2026-07 conditional novelty 6.0 of 10

    Rem3Di builds fixed-length, chirality-aware molecular descriptors by aggregating frozen atomistic foundation-model features with attention and a self-supervised denoising pretraining objective, matching or beating gra...

  5. Knowledge Distillation of a Protein Language Model Yields a Foundational Implicit Solvent Model

    physics.bio-ph 2026-01 conditional novelty 6.0 of 10

    A 45,000-parameter GNN distilled from a protein language model's secondary-structure predictions can drive MD simulations and, combined with GBn2 electrostatics, approximately reproduces explicit-solvent folding profi...

  6. Insights into CO dimerization at electrified Cu interfaces from large-scale machine learning simulations

    cond-mat.mtrl-sci 2025-09 reject novelty 6.0 of 10

    OC25 is a large open dataset and baseline models for solid-liquid interfaces, but the claimed CO dimerization insights are absent from the manuscript body.

  7. Towards Accurate and Scalable High-throughput MOF Adsorption Screening: Merging Classical Force Fields and Universal Machine Learned Interatomic Potentials

    cond-mat.mtrl-sci 2025-09 conditional novelty 6.0 of 10

    A hybrid MOF screening workflow combining UFF and the PFP machine-learned potential, validated against DFT, identifies seven promising MOFs for humidity-tolerant ethylene capture.

  8. Hot-Ham: an accurate and efficient E(3)-equivariant machine-learning electronic structures calculation framework

    physics.comp-ph 2025-09 conditional novelty 6.0 of 10

    Hot-Ham combines Gaunt tensor products with a local-coordinate SO(2) convolution to predict DFT Hamiltonians accurately and efficiently across several material classes.

  9. Universal Machine Learning Potentials under Pressure

    cond-mat.mtrl-sci 2025-08 conditional novelty 6.0 of 10

    Universal machine learning interatomic potentials systematically lose accuracy under pressure up to 150 GPa, and fine-tuning on high-pressure DFT data recovers most of the lost performance.

  10. Universal Machine Learning Potential for Systems with Reduced Dimensionality

    cond-mat.mtrl-sci 2025-08 conditional novelty 6.0 of 10

    Benchmarking 11 universal machine learning interatomic potentials on a new 40,000-structure, 0D-3D dataset shows energy and geometry errors grow as dimensionality falls, with eSEN the most transferable.

  11. Clay Edges Are Dynamic Proton-conducting Networks Modulated by Structure and pH

    physics.chem-ph 2025-08 conditional novelty 6.0 of 10

    Montmorillonite clay edges are amphoteric and conduct protons via direct and water-mediated hops, with rates and directions controlled by pH and magnesium-for-aluminum substitution.

  12. Distillation of atomistic foundation models across architectures and chemical domains

    physics.comp-ph 2025-06 conditional novelty 6.0 of 10

    Distillation of atomistic foundation models via synthetic data yields 10x-100x faster student potentials with near-teacher accuracy.

  13. Comparing classical and machine learning force fields for modeling deformation of solid sorbents relevant for direct air capture

    cond-mat.mtrl-sci 2025-06 conditional novelty 6.0 of 10

    No tested force field, classical or machine-learned, achieves the 0.1 eV accuracy target for adsorption energies in deformable MOFs relevant to direct air capture.

  14. chemtrain-deploy: A parallel and scalable framework for machine learning potentials in million-atom MD simulations

    physics.comp-ph 2025-06 conditional novelty 6.0 of 10

    A model-agnostic JAX-to-LAMMPS framework runs machine learning potentials in million-atom multi-GPU molecular dynamics with near-ideal strong and weak scaling.

  15. Machine Learning Interatomic Potentials: library for efficient training, model development and simulation of molecular systems

    physics.chem-ph 2025-05 conditional novelty 6.0 of 10

    InstaDeep's mlip library ports MACE, NequIP, and ViSNet to JAX with a JAX-MD backend, ships SPICE2-trained organics models, reports faster MD steps than its own Torch routes, and proposes a faster gated MACE variant i...

  16. Graph Neural Network Force Fields for Spin Dynamics in Metallic Magnets

    cond-mat.str-el 2026-07 conditional novelty 5.0 of 10

    A symmetry-aware GNN force field trained on s–d electronic data reproduces spin torques and nonequilibrium dynamics for collinear, coplanar, and noncoplanar metallic magnets.

  17. Fast and Accurate Foundation Models for Equivariant Machine-Learned Interatomic Potentials

    physics.comp-ph 2026-07 accept novelty 5.0 of 10

    Fast NequIP/Allegro foundation MLIPs reach leading MD inference speeds and strong benchmark accuracy; materials-discovery gains need better chemical diversity and consistent transition-metal energy surfaces.

  18. MANDALA: An E(3)-Equivariant Graph Neural Network Framework for Learning Electronic-Structure Operators with Observable Guidance

    cond-mat.mtrl-sci 2026-07 conditional novelty 5.0 of 10

    Mandala is a modular E(3)-equivariant GNN framework that learns sparse DFT Hamiltonian, overlap, and density matrices and supervises them with operator-derived observables.

  19. Full-data accuracy with fewer labels for training and fine-tuning machine-learning force fields

    physics.chem-ph 2026-07 conditional novelty 5.0 of 10

    Last-layer-projection-regression uncertainty lets active learning reach full-data accuracy with far fewer labels for both train-from-scratch and fine-tuned ML force fields, and can detect stable-but-wrong molecular dynamics.

  20. Benchmarking foundation potentials against quantum chemistry methods for predicting molecular redox potentials

    physics.chem-ph 2025-10 conditional novelty 5.0 of 10

    MACE-OMol foundation potential matches DFT accuracy for PCET redox potentials, but fails for multi-electron-transfer ions; a hybrid FP+DFT single-point workflow fixes this.

  21. ChemGraph: An Agentic Framework for Computational Chemistry Workflows

    physics.chem-ph 2025-06 conditional novelty 5.0 of 10

    A new LLM-driven framework, ChemGraph, automates molecular simulation workflows and shows that multi-agent task decomposition improves smaller models' accuracy on complex thermochemistry benchmarks.

Pith tools