E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate Interatomic Potentials

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.

References in corpus (31)

• PyTorch: An Imperative Style, High-Performance Deep Learning Library
• Decoupled Weight Decay Regularization
• Neural Message Passing for Quantum Chemistry
• Gaussian Approximation Potentials: the accuracy of quantum mechanics, without the electrons
• Gaussian Error Linear Units (GELUs)
• Deep Potential Molecular Dynamics: a scalable model with the accuracy of quantum mechanics
• ANI-1: An extensible neural network potential with DFT accuracy at force field computational cost
• E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate Interatomic Potentials
• On the Convergence of Adam and Beyond
• Moment Tensor Potentials: a class of systematically improvable interatomic potentials
• Quantum-Chemical Insights from Deep Tensor Neural Networks
• Machine Learning of Accurate Energy-Conserving Molecular Force Fields
• A Spectral Analysis Method for Automated Generation of Quantum-Accurate Interatomic Potentials
• PhysNet: A Neural Network for Predicting Energies, Forces, Dipole Moments and Partial Charges
• Towards Exact Molecular Dynamics Simulations with Machine-Learned Force Fields
• Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds
• SchNet: A continuous-filter convolutional neural network for modeling quantum interactions
• Deep Learning Scaling is Predictable, Empirically
• Ab initio thermodynamics of liquid and solid water
• Directional Message Passing for Molecular Graphs
• SpookyNet: Learning Force Fields with Electronic Degrees of Freedom and Nonlocal Effects
• FCHL revisited: faster and more accurate quantum machine learning
• Equivariant message passing for the prediction of tensorial properties and molecular spectra
• 3D Steerable CNNs: Learning Rotationally Equivariant Features in Volumetric Data
• End-to-end Symmetry Preserving Inter-atomic Potential Energy Model for Finite and Extended Systems
• Cormorant: Covariant Molecular Neural Networks
• GemNet: Universal Directional Graph Neural Networks for Molecules
• Isotope Effects in Liquid Water via Deep Potential Molecular Dynamics
• Fast Neural Network Approach for Direct Covariant Forces Prediction in Complex Multi-Element Extended Systems
• Accurate and scalable multi-element graph neural network force field and molecular dynamics with direct force architecture
• Geometric Prediction: Moving Beyond Scalars

Cited by in corpus (262)

• E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate Interatomic Potentials
• DeePMD-kit v2: A software package for Deep Potential models
• SpookyNet: Learning Force Fields with Electronic Degrees of Freedom and Nonlocal Effects
• A Comprehensive Survey on Deep Graph Representation Learning
• The Open Catalyst 2022 (OC22) Dataset and Challenges for Oxide Electrocatalysts
• Scalable Parallel Algorithm for Graph Neural Network Interatomic Potentials in Molecular Dynamics Simulations
• 14 Examples of How LLMs Can Transform Materials Science and Chemistry: A Reflection on a Large Language Model Hackathon
• Unifying the design space and optimizing linear and nonlinear truss metamaterials by generative modeling
• Inverse design of 3d molecular structures with conditional generative neural networks
• General-purpose machine-learned potential for 16 elemental metals and their alloys
• A practical guide to machine learning interatomic potentials -- Status and future
• Evaluation of the MACE Force Field Architecture: from Medicinal Chemistry to Materials Science
• Neural Network Potentials for Chemistry: Concepts, Applications and Prospects
• DPA-2: a large atomic model as a multi-task learner
• Ab-initio quantum chemistry with neural-network wavefunctions
• How to validate machine-learned interatomic potentials
• General framework for E(3)-equivariant neural network representation of density functional theory Hamiltonian
• Unified Graph Neural Network Force-field for the Periodic Table
• TeaNet: universal neural network interatomic potential inspired by iterative electronic relaxations
• Performance Assessment of Universal Machine Learning Interatomic Potentials: Challenges and Directions for Materials' Surfaces
• Current and future directions in network biology
• Advances of Machine Learning in Materials Science: Ideas and Techniques
• End-to-End Differentiable Molecular Mechanics Force Field Construction
• DeePMD-kit v3: A Multiple-Backend Framework for Machine Learning Potentials
• Informing Geometric Deep Learning with Electronic Interactions to Accelerate Quantum Chemistry
• Single-model uncertainty quantification in neural network potentials does not consistently outperform model ensembles
• SchNetPack 2.0: A neural network toolbox for atomistic machine learning
• Transferable E(3) equivariant parameterization for Hamiltonian of molecules and solids
• Tutorial: How to Train a Neural Network Potential
• Thermodynamics and dielectric response of ${\text{BaTiO}}_3$ by data-driven modeling
• Fast Uncertainty Estimates in Deep Learning Interatomic Potentials
• AI-driven inverse design of materials: Past, present and future
• Universal Machine Learning for the Response of Atomistic Systems to External Fields
• Deep learning generative model for crystal structure prediction
• Equivariant graph neural networks for fast electron density estimation of molecules, liquids, and solids
• i-PI 3.0: a flexible and efficient framework for advanced atomistic simulations
• The Evolution of Machine Learning Potentials for Molecules, Reactions and Materials
• Deep-learning electronic-structure calculation of magnetic superstructures
• Accelerating CALPHAD-based Phase Diagram Predictions in Complex Alloys Using Universal Machine Learning Potentials: Opportunities and Challenges
• Machine learning accelerated discovery of corrosion-resistant high-entropy alloys
• Deep-learning density functional perturbation theory
• TorchMD-Net 2.0: Fast Neural Network Potentials for Molecular Simulations
• Deep Ensembles vs. Committees for Uncertainty Estimation in Neural-Network Force Fields: Comparison and Application to Active Learning
• Inverse Design of Next-generation Superconductors Using Data-driven Deep Generative Models
• Neural functional theory for inhomogeneous fluids: Fundamentals and applications
• Spin-Dependent Graph Neural Network Potential for Magnetic Materials
• Uncertainty-biased molecular dynamics for learning uniformly accurate interatomic potentials
• Spectroscopy from Machine Learning by Accurately Representing the Atomic Polar Tensor
• High-accuracy thermodynamic properties to the melting point from ab initio calculations aided by machine-learning potentials
• Accurate machine learning force fields via experimental and simulation data fusion
• Deep Coarse-grained Potentials via Relative Entropy Minimization
• Force-Field-Enhanced Neural Network Interactions: from Local Equivariant Embedding to Atom-in-Molecule properties and long-range effects
• A reactive neural network framework for water-loaded acidic zeolites
• Point-particle drag, lift, and torque closure models using machine learning: hierarchical approach and interpretability
• Towards Linearly Scaling and Chemically Accurate Global Machine Learning Force Fields for Large Molecules
• Overcoming the Barrier of Orbital-Free Density Functional Theory for Molecular Systems Using Deep Learning
• Tensor-reduced atomic density representations
• Recent progress in the JARVIS infrastructure for next-generation data-driven materials design
• Machine Learning Potential for Electrochemical Interfaces with Hybrid Representation of Dielectric Response
• Machine Learning - Driven Materials Discovery: Unlocking Next-Generation Functional Materials - A review
• Layer-by-layer phase transformation in Ti${}_3$O${}_5$ revealed by machine learning molecular dynamics simulations
• An equivariant graph neural network for the elasticity tensors of all seven crystal systems
• E(n)-Equivariant Cartesian Tensor Passing Potential
• ArcaNN: automated enhanced sampling generation of training sets for chemically reactive machine learning interatomic potentials
• On the design space between molecular mechanics and machine learning force fields
• Robustness of Local Predictions in Atomistic Machine Learning Models
• Advances in modeling complex materials: The rise of neuroevolution potentials
• Artificial Intelligence in Materials Science and Engineering: Current Landscape, Key Challenges, and Future Trajectorie
• Multilayer atomic cluster expansion for semi-local interactions
• Model-free quantification of completeness, uncertainties, and outliers in atomistic machine learning using information theory
• Grappa -- A Machine Learned Molecular Mechanics Force Field
• High Pressure and Temperature Neural Network Reactive Force Field for Energetic Materials
• Temporally coherent backmapping of molecular trajectories from coarse-grained to atomistic resolution
• Lorentz group equivariant autoencoders
• Machine Learning and Data-Driven Methods in Computational Surface and Interface Science
• Completeness of Atomic Structure Representations
• Generative diffusion model for surface structure discovery
• Denoise Pretraining on Nonequilibrium Molecules for Accurate and Transferable Neural Potentials
• Heat flux for semi-local machine-learning potentials
• Lightweight and Effective Tensor Sensitivity for Atomistic Neural Networks
• AI-Driven Accelerated Discovery of Intercalation-type Cathode Materials for Magnesium Batteries
• Accelerated chemical space search using a quantum-inspired cluster expansion approach
• Machine Learned Interatomic Potential for Dispersion Strengthened Plasma Facing Components
• Reinforcement learning for traversing chemical structure space: Optimizing transition states and minimum energy paths of molecules
• Accelerating the prediction of inorganic surfaces with machine learning interatomic potentials
• NepoIP/MM: Towards Accurate Biomolecular Simulation with a Machine Learning/Molecular Mechanics Model Incorporating Polarization Effects
• Geometry-enhanced Pre-training on Interatomic Potentials
• Neural Network Potential with Multi-Resolution Approach Enables Accurate Prediction of Reaction Free Energies in Solution
• Why neural functionals suit statistical mechanics
• 3DReact: Geometric deep learning for chemical reactions
• Accelerating discrete dislocation dynamics simulations with graph neural networks
• Kernel based quantum machine learning at record rate : Many-body distribution functionals as compact representations
• Surrogate- and invariance-boosted contrastive learning for data-scarce applications in science
• Machine-Learning Interatomic Potentials for Long-Range Systems
• Data efficiency and extrapolation trends in neural network interatomic potentials
• chemtrain: Learning Deep Potential Models via Automatic Differentiation and Statistical Physics
• Ab Initio Generalized Langevin Equation
• MGNN: Moment Graph Neural Network for Universal Molecular Potentials
• Constructing and evaluating machine-learned interatomic potentials for Li-based disordered rocksalts
• Accurate quantum Monte Carlo forces for machine-learned force fields: Ethanol as a benchmark
• Reducing the cost of neural network potential generation for reactive molecular systems
• Accelerating QM/MM simulations of electrochemical interfaces through machine learning of electronic charge densities
• Stress and heat flux via automatic differentiation
• Intramolecular and water mediated tautomerism of solvated glycine
• Optimal pre-train/fine-tune strategies for accurate material property predictions
• Artificial Intelligence for Science in Quantum, Atomistic, and Continuum Systems
• Predicting ionic conductivity in solids from the machine-learned potential energy landscape
• Analyzing Atomic Interactions in Molecules as Learned by Neural Networks
• Unified Deep Learning Framework for Many-Body Quantum Chemistry via Green's Functions
• Advancing Molecular Machine Learning Representations with Stereoelectronics-Infused Molecular Graphs
• FeNNol: an Efficient and Flexible Library for Building Force-field-enhanced Neural Network Potentials
• Machine-Learned Atomic Cluster Expansion Potentials for Fast and Quantum-Accurate Thermal Simulations of Wurtzite AlN
• GPU-Accelerated Approximate Kernel Method for Quantum Machine Learning
• Approximately Equivariant Quantum Neural Network for $p4m$ Group Symmetries in Images
• Random sampling versus active learning algorithms for machine learning potentials of quantum liquid water
• A collinear-spin machine learned interatomic potential for Fe${}_7$Cr${}_2$Ni alloy
• Exploring Li-ion Transport Properties of Li${}_3$TiCl${}_6$: A Machine Learning Molecular Dynamics Study
• Augmentation of Universal Potentials for Broad Applications
• Exploring the design space of machine-learning models for quantum chemistry with a fully differentiable framework
• Modelling complex proton transport phenomena -- Exploring the limits of fine-tuning and transferability of foundational machine-learned force fields
• A New Moiré Platform Based on M-Point Twisting
• Atomistic modeling of bulk and grain boundary diffusion in solid electrolyte Li${}_6$PS${}_5$Cl using machine-learning interatomic potentials
• How Graph Neural Network Interatomic Potentials Extrapolate: Role of the Message-Passing Algorithm
• Beyond MD17: the reactive xxMD dataset
• Transferable empirical pseudopotenials from machine learning
• Rotation-equivariant Graph Neural Networks for Learning Glassy Liquids Representations
• Supervised Learning and the Finite-Temperature String Method for Computing Committor Functions and Reaction Rates
• Structure discovery in Atomic Force Microscopy imaging of ice
• Probing the effects of broken symmetries in machine learning
• Machine Learning Models Capture Plasmon Dynamics in Ag Nanoparticles
• Machine learning and invariant theory
• General time-reversal equivariant neural network potential for magnetic materials
• Direct imaging of carbohydrate stereochemistry
• Cascading symmetry constraint during machine learning-enabled structural search for sulfur induced Cu(111)-$(\sqrt{43}\times \sqrt{43})$ surface reconstruction
• DPmoire: A tool for constructing accurate machine learning force fields in moiré systems
• Neural Thermodynamic Integration: Free Energies from Energy-based Diffusion Models
• Size and Quality of Quantum Mechanical Data Sets for Training Neural Network Force Fields for Liquid Water
• Efficient mapping of phase diagrams with conditional Boltzmann Generators
• Towards Symbolic XAI -- Explanation Through Human Understandable Logical Relationships Between Features
• Machine learning for structure-property relationships: Scalability and limitations
• Metatensor and metatomic: foundational libraries for interoperable atomistic machine learning
• Exploration of amorphous V${}_2$O${}_5$ as cathode for magnesium batteries
• Iterative charge equilibration for fourth-generation high-dimensional neural network potentials
• Extending the atomic decomposition and many-body representation, a chemistry-motivated monomer-centered approach for machine learning potentials
• Beyond Euclid: An Illustrated Guide to Modern Machine Learning with Geometric, Topological, and Algebraic Structures
• Does Hessian Data Improve the Performance of Machine Learning Potentials?
• Fitting to magnetic forces improves the reliability of magnetic Moment Tensor Potentials
• A smooth basis for atomistic machine learning
• Equivariant neural network for Green's functions of molecules and materials
• Automated atomistic simulations of dissociated dislocations with ab initio accuracy
• $\Delta$-model correction of Foundation Model based on the models own understanding
• Beam induced heating in electron microscopy modeled with machine learning interatomic potentials
• Fast proper orthogonal descriptors for many-body interatomic potentials
• Equivariant Neural Networks for Spin Dynamics Simulations of Itinerant Magnets
• Machine learning frontier orbital energies of nanodiamonds
• AMARO: All Heavy-Atom Transferable Neural Network Potentials of Protein Thermodynamics
• Efficient moment tensor machine-learning interatomic potential for accurate description of defects in Ni-Al Alloys
• Graph-neural-network predictions of solid-state NMR parameters from spherical tensor decomposition
• Investigating Ionic Diffusivity in Amorphous Solid Electrolytes using Machine Learned Interatomic Potentials
• Efficient ensemble uncertainty estimation in Gaussian Processes Regression
• Phase stability in the Hf-N and Zr-N systems
• Study of vacancy ordering and the boson peak in metastable cubic Ge-Sb-Te using machine learning potentials
• A Comprehensive Assessment and Benchmark Study of Large Atomistic Foundation Models for Phonons
• Efficient Generation of Stable Linear Machine-Learning Force Fields with Uncertainty-Aware Active Learning
• Equivariant Machine Learning Interatomic Potentials with Global Charge Redistribution
• Scaling Graph Neural Networks to Large Proteins
• On-the-fly machine learning-augmented constrained AIMD to design new routes from glassy carbon to quenchable amorphous diamond with low pressure and temperature
• Scaling Molecular Dynamics with ab initio Accuracy to 149 Nanoseconds per Day
• Molecular relaxation by reverse diffusion with time step prediction
• A neural network potential with self-trained atomic fingerprints: a test with the mW water potential
• EOSnet: Embedded Overlap Structures for Graph Neural Networks in Predicting Material Properties
• Efficient dataset generation for machine learning perovskite alloys
• Compressing and forecasting atomic material simulations with descriptors
• Crystal structure prediction with host-guided inpainting generation and foundation potentials
• Kapitza stabilization of quantum critical order
• Reconstructions and Dynamics of $\beta$-Lithium Thiophosphate Surfaces
• Apax: A Flexible and Performant Framework For The Development of Machine-Learned Interatomic Potentials
• Superconductor discovery in the emerging paradigm of Materials Informatics
• Atomistic Simulations of Oxide-Water Interfaces using Machine Learning Potentials
• Multi-head committees enable direct uncertainty prediction for atomistic foundation models
• Rapidly Encoding Generalizable Dynamics in a Euclidean Symmetric Neural Network
• Modeling Chemical Exfoliation of Non-van der Waals Chromium Sulfides by Machine Learning Interatomic Potentials and Monte Carlo Simulations
• Accelerating structure search using atomistic graph-based classifiers
• A Planning-and-Exploring Approach to Extreme-Mechanics Force Fields
• Hydrogen liquid-liquid transition from first principles and machine learning
• Accelerated lignocellulosic molecule adsorption structure determination
• Trade-off between Gradient Measurement Efficiency and Expressivity in Deep Quantum Neural Networks
• Equivariant Tensor Network Potentials
• Energy-GNoME: A Living Database of Selected Materials for Energy Applications
• Application-specific machine-learned interatomic potentials: exploring the trade-off between DFT convergence, MLIP expressivity, and computational cost
• Identifying Split Vacancy Defects with Machine-Learned Foundation Models and Electrostatics
• Thermal Conductivity Calculation using Homogeneous Non-equilibrium Molecular Dynamics Simulation with Allegro
• Evaluating the Transferability of Machine-Learned Force Fields for Material Property Modeling
• High-Accuracy Physical Property Prediction for Organics via Molecular Representation Learning: Bridging Data to Discovery
• Fine-Tuning Unifies Foundational Machine-learned Interatomic Potential Architectures at ab initio Accuracy
• HEroBM: a deep equivariant graph neural network for universal backmapping from coarse-grained to all-atom representations
• Unveiling the Lithium-Ion Transport Mechanism in Li2ZrCl6 Solid-State Electrolyte via Deep Learning-Accelerated Molecular Dynamics Simulations
• Deriving effective electrode-ion interactions from free-energy profiles at electrochemical interfaces
• Machine learning interatomic potential for the low-modulus Ti-Nb-Zr alloys in the vicinity of dynamical instability
• Self-learning Monte Carlo with equivariant Transformer
• Automated Machine Learning Pipeline: Large Language Models-Assisted Automated Dataset Generation for Training Machine-Learned Interatomic Potentials
• Equivariant Electronic Hamiltonian Prediction with Many-Body Message Passing
• Free energy profiles for chemical reactions in solution from high-dimensional neural network potentials: The case of the Strecker synthesis
• Tokenization for Molecular Foundation Models
• Noise Tolerant Force Calculations in Density Functional Theory: A Surface Integral Approach for Wavelet-Based Methods
• Intermediates of Forming Transition Metal Dichalcogenide Heterostructures Revealed by Machine Learning Simulations
• Strength of 2D Glasses Explored by Machine-Learning Force Fields
• Zero Shot Molecular Generation via Similarity Kernels
• Teaching oxidation states to neural networks
• Data Fusion of Deep Learned Molecular Embeddings for Property Prediction
• A Data-Driven Approach to Coarse-Graining Simple Liquids in Confinement
• Energy-Based Coarse-Graining in Molecular Dynamics: A Flow-Based Framework without Data
• Accurate Machine Learning Interatomic Potentials for Polyacene Molecular Crystals: Application to Single Molecule Host-Guest Systems
• Modeling phase transformations in Mn-rich disordered rocksalt cathodes with machine learning interatomic potentials
• A Generative Diffusion Model for Amorphous Materials
• Extreme anharmonicity and thermal contraction of 1D wires
• Low-rank matrix and tensor approximations for compression of machine-learning interatomic potentials
• Learning atomic forces from uncertainty-calibrated adversarial attacks
• Towards Improved Quantum Machine Learning for Molecular Force Fields
• "One defect, one potential" strategy for accurate machine learning prediction of defect phonons
• Modular hybrid machine learning and physics-based potentials for scalable modeling of van der Waals heterostructures
• Considerations in the use of ML interaction potentials for free energy calculations
• FIRE-GNN: Force-informed, Relaxed Equivariance Graph Neural Network for Rapid and Accurate Prediction of Surface Properties
• Reinforcement Fine-Tuning for Materials Design
• Self-supervised Representations and Node Embedding Graph Neural Networks for Accurate and Multi-scale Analysis of Materials
• Probing the Temporal Response of Liquid Water to a THz Pump Pulse Using Machine Learning-Accelerated Non-Equilibrium Molecular Dynamics
• FastCHGNet: Training one Universal Interatomic Potential to 1.5 Hours with 32 GPUs
• Reconstructing Kernel-based Machine Learning Force Fields with Super-linear Convergence
• Deep-learning atomistic semi-empirical pseudopotential model for nanomaterials
• Machine-Learned Potentials for Solvation Modeling
• Efficient Learning of Long-Range and Equivariant Quantum Systems
• Machine-Learned Bond-Order Potential for Exploring the Configuration Space of Carbon
• Artificial Intelligence for Direct Prediction of Molecular Dynamics Across Chemical Space
• General Learning of the Electric Response of Inorganic Materials
• Active Δ-learning with universal potentials for global structure optimization
• Crystal Nucleation in Eutectic Al-Si Alloys by Machine-Learned Molecular Dynamics
• Representing spherical tensors with scalar-based machine-learning models
• Cation Dominated but Negatively Charged Na2SO4,aq-Graphene Interfaces
• Multi-objective optimization and quantum hybridization of equivariant deep learning interatomic potentials
• Machine-learning force-field models for dynamical simulations of metallic magnets
• Multi-resolution Physics-Aware Recurrent Convolutional Neural Network for Complex Flows
• Self-consistency error correction for accurate machine learning potentials from variational Monte Carlo
• Convergence of Body-Orders in Linear Atomic Cluster Expansions
• Decoding lithium's subtle phase stability with a machine learning force field
• Exact average many-body interatomic interaction model for random alloys
• Scalable Training of Neural Network Potentials for Complex Interfaces Through Data Augmentation
• XMCQDPT2-Fidelity Transfer-Learning Potentials and a Wavepacket Oscillation Model with Power-Law Decay for Ultrafast Photodynamics
• Gradient-based grand canonical optimization enabled by graph neural networks with fractional atomic existence
• Assessing zero-shot generalisation behaviour in graph-neural-network interatomic potentials
• Algebras of actions in an agent's representations of the world
• Compressing local atomic neighbourhood descriptors
• Active Learning of a Neural Network Potential for Large-Scale Atomistic Simulations of Polymer Electrolyte Membranes
• Neural Polarization: Toward Electron Density for Molecules by Extending Equivariant Networks
• Knowledge Distillation of a Protein Language Model Yields a Foundational Implicit Solvent Model
• Atoms to Processes: The Role of Artificial Intelligence and Machine Learning in Chemical Engineering
• Machine learning of electronic structure and atomistic properties from the external potential
• Score-based diffusion models for accurate crystal-structure inpainting and reconstruction of hydrogen positions
• GEWUM: General Exploration Workflow for the Utopia of Materials: A Unified Platform for Automated Structure Generation, Selection, and Validation
• Irregular Metamaterial Networks
• A flexible kinetic Monte Carlo framework for GaN molecular beam epitaxy with adaptive on-the-fly barrier evaluation
• Inverse design of bespoke interatomic potentials via active learning by information-matching
• Machine-learning modeling of magnetization dynamics in quasi-equilibrium and driven metallic spin systems