REVIEW 43 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.
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
MACE: Higher Order Equivariant Message Passing Neural Networks for Fast and Accurate Force Fields
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.
Forward citations
Cited by 43 Pith papers
-
A fast summation method for the DFT-D3 dispersion correction
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.
-
Adaptive fine-tuning of foundation models for crystal structure prediction: Discovery of high-pressure phases in the CaFeNi system
An adaptive fine-tuning workflow for foundation-model MLIPs enables efficient CSP in the CaFeNi ternary, reproducing the low-pressure hull and predicting a new phase Ca6FeNi stable above 100 GPa.
-
A Unified Framework for Vision Transformers Equivariant to Discrete Subgroups of $\mathrm{O}(2)$
A unified family of vision transformers equivariant to arbitrary discrete subgroups of O(2), with embedding and expressivity theorems, a D6 construction using hexagonal patches, and experiments on aerial images in low...
-
A Unified Framework for Vision Transformers Equivariant to Discrete Subgroups of $\mathrm{O}(2)$
Constructs G-equivariant ViTs for arbitrary discrete G ≤ O(2), proves H ≤ G implies G-models embed into H-models and single-head equivariant attention realizes all ordinary G-equivariant maps, introduces D6 hexagonal ...
-
Constraint-Aware Quantum Optimization of Defect Configurations in Doped ZrO2: XY-Mixer QAOA and Grover Adaptive Search
Presents an end-to-end constraint-aware quantum optimization pipeline using XY-mixer QAOA and Grover Adaptive Search for low-energy defect configurations in doped ZrO2, with QAOA validated against exact enumeration on...
-
Magnetic HIP-NN for spin dynamics in disordered itinerant magnets
mHIP-NN extends hierarchical message-passing networks to model electron-mediated spin dynamics in disordered itinerant magnets while preserving rotational symmetry.
-
Disorder-driven symmetry suppression by van der Waals planar defects in a magnetic topological insulator
Ion irradiation creates a layer-disordered phase with planar defects in MnBi2Te4 that suppresses symmetry and anomalous Hall conductivity fivefold while magnetic order persists.
-
Fast contracted Clebsch--Gordan tensor products for equivariant graph neural networks
An O(L^3) algorithm computes contracted Clebsch-Gordan tensor products for equivariant ML potentials using a structured angular grid and spherical Poisson bracket to handle parity-odd terms at fixed CP rank.
-
Force-Aware Neural Tangent Kernels for Scalable and Robust Active Learning of MLIPs
Force-aware NTKs and chunked acquisition enable scalable, robust active learning for MLIPs, achieving lowest energy and force errors on OC20 and remaining competitive on other benchmarks.
-
QT-Net: Rethinking Evaluation of AI Models in Atomic Chemical Space
QT-Net predicts atomic electron populations and multipoles via a new SOAP-cluster held-out test, improving molecular property prediction and recovering QM9 dipole moments from per-atom outputs.
-
Pretrained Model Representations as Acquisition Signals for Active Learning of MLIPs
Pretrained MLIP latent spaces yield NTK and activation kernels that outperform standard acquisition functions in active learning for reactive MLIP training, reducing required labels by 38% for energy and 28% for force errors.
-
Pretrained Model Representations as Acquisition Signals for Active Learning of MLIPs
Kernels from pretrained MLIP latent spaces outperform standard acquisition methods in active learning for reactive chemistry, reducing required labels by 38% for energy error and 28% for force error.
-
Graph Neural Networks in the Wilson Loop Representation of Abelian Lattice Gauge Theories
A gauge-invariant GNN using Wilson loops as inputs accurately predicts observables and simulates dynamics in Z2 and U(1) lattice gauge models.
-
DeepHartree: A Poisson-Coupled Neural Field for One-Shot Density Functional Theory
A Poisson-coupled equivariant neural field predicts mutually consistent densities and potentials to accelerate DFT SCF convergence by up to 40.6% with zero-shot transfer to larger systems.
-
DeepHartree: A Poisson-Coupled Neural Field for One-Shot Density Functional Theory
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.
-
Enabling Biomolecular Simulations with Neural Network Potentials in GROMACS
A flexible interface in GROMACS integrates PyTorch neural network potentials into hybrid ML/MM molecular dynamics simulations for biomolecular systems.
-
Rem3Di: Learning smooth, chiral 3D molecular descriptors from atomistic foundation models
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...
-
Property-Specific Molecular Representations via Feature-Space Transfer Compression
A transfer compression technique using semi-empirical data reduces molecular representation dimensions by a median 72% (range 36-98%) while retaining accuracy for energy, heat capacity, dipole moment and polarizabilit...
-
Computational references are not experiments: pre-registered validation of machine-learned sodium-cathode voltages
Pre-registered validation of an ML Na-cathode voltage screen yields 0.67 V MAE against experiment, with Materials Project PBE+U references 0.54 V low and voltage-dependent residuals preventing calibration, leading to ...
-
Computational references are not experiments: pre-registered validation of machine-learned sodium-cathode voltages
Pre-registered validation of an ML Na-cathode voltage screen yields 0.67 V MAE against experiment, with Materials Project PBE+U references 0.54 V low and dominating the error.
-
Multitask learning with semiempirical orbital charges enables sample-efficient MLIPs
Multitask learning on linear-scaling GFN1-xTB orbital charges cuts energy MAE by 46% and data needs by 5x versus energy-only MLIPs while outperforming DFT atomic charge augmentation.
-
Anisotropic Crystallization Kinetics and Interfacial Dynamics of Phase-Change Material Sb$_2$S$_3$ from Machine Learning Force Field Simulations
Machine learning force field molecular dynamics simulations reveal anisotropic crystallization in Sb2S3 with [100] facet fastest growth and interface-controlled kinetics with activation energy 0.55-0.57 eV.
-
Reweighting free energy profiles between universal machine learning interatomic potentials for fast consensus building
A reweighting method with mean energy-gap approximation transfers PMFs between MLIPs to recover target reaction and activation free energies at low cost for a 601-atom Li+ transport system across DFT levels.
-
Force-Aware Neural Tangent Kernels for Scalable and Robust Active Learning of MLIPs
Force-aware Neural Tangent Kernels combined with chunked acquisition provide scalable and distribution-robust active learning for MLIPs, outperforming baselines on OC20 and remaining competitive on other benchmarks.
-
Fast and Accurate Prediction of Lattice Thermal Conductivity via Machine Learning Surrogates
Machine learning models, especially certain deep neural networks, can predict lattice thermal conductivity with useful accuracy across different generalization tests while being orders of magnitude faster than first-p...
-
Towards Accelerated SCF Workflows with Equivariant Density-Matrix Learning and Analytic Refinement
dm-PhiSNet predicts 1-RDMs from geometries via equivariant PhiSNet with two-stage training and analytic refinement, reducing SCF iterations 49-81% on six closed-shell molecules while giving accurate one-shot energies ...
-
Data-Driven Thermal and Mechanical Modeling of Defective Covalent Organic Frameworks
QCOF ML potentials tuned on COF data outperform general MACE models for defective systems and reveal higher thermal defect sensitivity in CTF-1 versus COF-LZU1 with nearly invariant low-strain mechanics.
-
Force Field-Agnostic Phase Classification of Zeolitic Imidazolate Framework Polymorphs
Neural networks trained on molecular configurations from different force fields classify ZIF polymorph phases accurately in simulations and expose transition mechanisms without force-field bias.
-
QUASAR: A Universal Autonomous System for Atomistic Simulation and a Benchmark of Its Capabilities
QUASAR is a new autonomous LLM-based system that orchestrates multi-scale atomistic simulations and benchmarks as a general reasoning tool rather than a narrow automation script.
-
Bayesian E(3)-Equivariant Interatomic Potential with Iterative Restratification of Many-body Message Passing
Bayesian E(3)-equivariant MLPs with joint energy-force NLL loss achieve competitive accuracy while enabling uncertainty-guided active learning, OOD detection, and calibration.
-
Reducing Self-Interaction Error in Transition-Metal Oxides with Different Exact-Exchange Fractions for Energy and Density
r²SCANY@r²SCANX uses distinct exact-exchange fractions for density (X) and energy (Y) to reduce self-interaction errors and improve band gaps, magnetic moments, and oxidation energies in 20 strongly correlated transit...
-
Graph Neural Network Force Fields for Spin Dynamics in Metallic Magnets
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.
-
Fast and Accurate Foundation Models for Equivariant Machine-Learned Interatomic Potentials
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.
-
MANDALA: An E(3)-Equivariant Graph Neural Network Framework for Learning Electronic-Structure Operators with Observable Guidance
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.
-
Full-data accuracy with fewer labels for training and fine-tuning machine-learning force fields
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.
-
A Hybrid Quantum Mechanics Machine Learning Forcefield (QM/ML) Framework for Accurate Solute-Dislocation Interaction Simulations
Hybrid QM/ML forcefield framework couples DFT with MLIPs to enable scalable, chemically accurate simulations of solute-dislocation interactions, demonstrated on Sn/Fe segregation in Zr and magnetic effects in steel.
-
Loss-Guided Adaptive Scale Refinement for Molecular Force Prediction
Loss-guided adaptive scale refinement on NaCl aqueous system reduces overall force MAE from 399.65 to 381.23 by discovering intermediate scales from initial anchors.
-
Disorder-driven symmetry suppression by van der Waals planar defects in a magnetic topological insulator
Ion irradiation creates van der Waals planar defects in MnBi2Te4 that reduce magnetic anisotropy and suppress anomalous Hall effect by over an order of magnitude while preserving antiferromagnetic order and partial cr...
-
Electron dynamics mediate the water-carbon {\pi} bond
Electron dynamics in the aromatic pi cloud quench and amplify specific vibrational signals from a single water molecule bound to pyrene anion.
-
Machine-Learned Force Fields for Lattice Dynamics at Coupled-Cluster Level Accuracy
Machine-learned force fields trained on coupled-cluster potential energy surfaces produce phonon dispersions and vibrational densities of states for solids that agree better with experiment than DFT-based models.
-
Machine Learning Interatomic Potentials: Advancing Open-Source Software for Efficient and Scalable Molecular Simulation
mlip v2 is a new software release that integrates API redesign, e3j backend, eSEN model, improved charge modeling, and expanded simulation capabilities to support larger-scale molecular modeling.
-
Comparing fine-tuning strategies of MACE machine learning force field for modeling Li-ion diffusion in LiF for batteries
MACE-MPA-0 predicts Li diffusion Ea of 0.22 eV in LiF, fine-tuned version with 300 points gives 0.20 eV, close to DeePMD reference of 0.24 eV, using far less training data.
-
Accelerating point defect simulations using data-driven and machine learning approaches
Machine learning models trained on quantum mechanical data can predict defect properties in solids with high accuracy but at much lower computational cost than traditional methods.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.