Machine-Learning Interatomic Potentials for Molecular Simulation

Topic: machine-learning interatomic potentials for molecular simulation · Since 1980 · Grounded citations only · Published 2026-08-09

The core problem: you need the potential-energy surface

Almost everything atomistic simulation computes — molecular dynamics trajectories, free energies, phase diagrams, reaction rates, spectra — is downstream of one object: the potential-energy surface (PES), the total energy E({R}) as a function of every atomic position, together with the forces F_i = −∂E/∂R_i that follow from it. Get the PES right and the dynamics follow. Get it wrong and no amount of sampling helps.

There have historically been two ways to obtain it, and they fail in opposite directions.

Quantum-mechanical methods solve for the electronic structure at each configuration. Density-functional theory is the workhorse, implemented in mature plane-wave codes [1] [2], and it is accurate enough to be treated as reference data across chemistry and materials science. Its problem is cost: the scaling with system size makes long trajectories of large systems infeasible, which is exactly the statement that motivated the first neural-network PES work — "the accurate description of chemical processes often requires the use of computationally demanding methods like density-functional theory, making long simulations of large systems unfeasible" [3].

Classical force fields replace the electrons with a fixed functional form — bonded terms, Lennard-Jones, point charges — fitted to a limited set of data. They are cheap enough for million-atom, microsecond simulations in engines built for exactly that [4] [5] [6], and they underpin a great deal of applied work including pharmaceutical modelling [7]. Their problem is the fixed functional form: accuracy is bounded by what the form can express, transferability outside the fitting regime is poor, and bond breaking requires a specially constructed reactive form such as REBO or ReaxFF [8] [9].

Machine-learning interatomic potentials (MLIPs) target the gap. They fit the PES directly from quantum reference data using a flexible regressor with no fixed functional form, aiming for near-DFT accuracy at near-classical cost. The two founding papers state the bargain precisely: a neural-network representation of a DFT PES that "provides the energy and forces as a function of all atomic positions in systems of arbitrary size and is several orders of magnitude faster than DFT" [3]; and a kernel model whose "models do not have a fixed functional form and hence are capable of modeling complex potential energy landscapes… systematically improvable with more data," saving "orders of magnitude in computational cost" on the long trajectories needed for finite-temperature properties [10].

Two structural constraints shape every architecture that followed. First, the energy must be extensive and size-transferable, which is achieved by decomposing the total energy into a sum of atomic contributions, each a function of that atom's local environment within a cutoff [3] [11]. Second, physical symmetry must be respected: the energy is invariant under translation, rotation, and permutation of identical atoms, while forces rotate with the system. Whether a model imposes these by construction or has to learn them from data turns out to be the single largest determinant of data efficiency.

Method lineage

Family 1: descriptor-based neural networks

The Behler–Parrinello construction encodes each atom's neighbourhood in hand-designed symmetry functions — radial and angular features that are invariant by construction — and feeds them to a small per-element neural network whose outputs sum to the total energy [3]. The approach was demonstrated on bulk silicon against both empirical potentials and DFT, and the review a few years later documented the transition from low-dimensional to genuinely high-dimensional systems, with MD of thousands of atoms and energies and forces "obtained with very high accuracy," while being candid that development "requires substantial computational effort as typically thousands of reference calculations are required" [11].

The known limitation of this family is locality. Because each atomic energy depends only on a finite cutoff sphere, long-range electrostatics and global charge transfer are invisible. Fourth-generation high-dimensional NNPs address this with a charge-equilibration scheme built on environment-dependent atomic electronegativities, correctly describing global charge distributions in systems that purely local models get wrong [12].

Family 2: kernel methods and Gaussian processes

Gaussian Approximation Potentials take the non-parametric route: rather than a network, a Gaussian-process regressor over a smooth representation of the atomic environment, systematically improvable as data accumulates [10]. The statistical machinery is standard GP regression specialized to atomistic data [13], and the practical appeal is that a GP supplies a predictive variance alongside its prediction — an uncertainty estimate for free, which turns out to matter for active learning. GAP has been benchmarked head-to-head against permutationally invariant polynomials and neural networks on the same PES-fitting tasks [14] and used to build general-purpose potentials for hard cases such as silicon across its phase diagram [15].

Family 3: systematic linear expansions

A parallel line asks for a complete, systematically convergent basis for the local environment rather than a black-box regressor. Moment tensor potentials pose it as a mathematical approximation problem and construct a class of potentials that is systematically improvable by design [16]. The spectral neighbor analysis method (SNAP) expands the neighbour density in bispectrum components with linear fitting [17] [18]. The atomic cluster expansion (ACE) generalizes these into a complete descriptor whose evaluation cost scales linearly with the number of neighbours irrespective of expansion order, converging smoothly to meV accuracy and — once nonlinear functions of the expansion are introduced — reaching accuracy comparable to state-of-the-art ML potentials [19]. ACE now has performant production implementations [20] [21] [22]. The attraction of this family is interpretability and control: the approximation error can be reasoned about rather than only measured.

Family 4: message-passing and graph neural networks

Hand-designed descriptors cap what a model can see. Graph neural networks instead learn the representation, treating atoms as nodes and neighbours as edges and iterating learned messages. SchNet introduced continuous-filter convolutional layers appropriate to atoms at arbitrary positions rather than on a grid, predicting properties across chemical space and producing energy-conserving force fields for MD [23] [24] [25]. Because plain distance-based messages discard angular information, follow-up architectures reintroduce it — directional message passing over molecular graphs [26] [27], and Newtonian-mechanics-inspired message construction [28].

Family 5: E(3)-equivariance — the step change

The decisive idea is that symmetry should be a property of the architecture, not something learned. Most earlier symmetry-aware models are invariant: they act only on scalars, which is correct for energy but throws away the directional structure of the environment. NequIP instead uses E(3)-equivariant convolutions acting on geometric tensors, so internal features rotate with the system and only the final energy is invariant — "a more information-rich and faithful representation of atomic environments" [29]. The empirical consequence is the headline result of the field: state-of-the-art accuracy "with up to three orders of magnitude fewer training data, challenging the widely held belief that deep neural networks require massive training sets," data-efficient enough to fit potentials to high-order quantum-chemical reference [29]. Symmetry is, in effect, a very strong prior — and priors buy data.

Two directions extend it. MACE attacks the cost side: standard MPNNs pass only two-body messages, tying expressivity to depth and therefore to communication cost. Using higher body-order messages — four-body — cuts the required message-passing iterations to two, giving a fast, highly parallelizable model at or beyond state-of-the-art accuracy on rMD17, 3BPA and AcAc [30], subsequently evaluated across amorphous carbon, universal materials modelling, small-molecule chemistry, large molecules and liquid water, and shown to be data-efficient enough to reproduce experimental vibrational spectra when trained on as few as 50 reference configurations [31]. Allegro attacks the scaling side: atom-centred message passing limits accessible length scales, so it builds a strictly local equivariant model from iterated tensor products of learned equivariant representations with no atom-centred message passing, keeping accuracy while scaling to large simulations [32]. Related equivariant designs cover tensorial properties and spectra [33], attention/transformer formulations [34] [35] [36], and the design space itself has been mapped as a unified framework [37] [38].

Family 6: transferable potentials across chemical space

The families above are typically fitted per-system. ANI targeted breadth instead: modified Behler–Parrinello symmetry functions built into atomic environment vectors, trained on data spanning both configurational and conformational space, giving "an accurate and transferable potential for organic molecules" — DFT accuracy at force-field cost [39]. That programme is fundamentally a data programme: 20 million off-equilibrium conformations for ANI-1 [40], then coupled-cluster and DFT property sets [41]. Its most instructive result is methodological — transfer learning from a large DFT-trained model onto a smaller gold-standard CCSD(T)/CBS set yields ANI-1ccx, approaching coupled-cluster accuracy on reaction thermochemistry, isomerization and drug-like torsions while remaining "billions of times faster" [42]. Accuracy ceiling and data volume can be decoupled. The same recipe now runs on equivariant backbones for organic chemistry [43] [44], supported by purpose-built datasets such as SPICE, over 1.1 million conformations of drug-like molecules, dimers, dipeptides and solvated amino acids at ωB97M-D3(BJ)/def2-TZVPPD [45].

Family 7: universal / foundation potentials

The current frontier extends transferability from organic chemistry to the whole periodic table, trained on the accumulated output of high-throughput DFT [46] [47]. M3GNet provides a universal graph deep-learning interatomic potential for the periodic table [48]; a universal NNP covering arbitrary combinations of 45 elements followed [49]; and CHGNet is pretrained on the Materials Project Trajectory Dataset — energies, forces, stresses and magnetic moments from over a decade of DFT on more than 1.5 million inorganic structures — with magnetic moments included explicitly so the model can represent electronic state alongside ionic rearrangement [50]. The emerging usage pattern mirrors the rest of machine learning: pretrain broadly, then fine-tune on a small system-specific set for quantitative accuracy [51].

A minimal model: the diatomic potential-energy curve

Before the high-dimensional machinery, it is worth being concrete about what a "potential" is, because the one-dimensional case contains most of the intuition.

For a diatomic molecule the PES collapses to a single coordinate, the internuclear separation r, and the standard closed form is the Morse potential:

V(r) = D_e · (1 − e^(−a(r − r_e)))²

with three parameters: r_e, the equilibrium bond length where the curve is at its minimum; D_e, the well depth, i.e. the dissociation energy measured from the bottom of the well; and a, which sets the curvature and therefore the vibrational frequency. Its two essential features are that it is anharmonic — asymmetric about the minimum, steeply repulsive at short range from nuclear and Pauli repulsion, and flattening to a finite asymptote as r → ∞ because the bond simply breaks — and that it is smooth and differentiable, so forces come from −dV/dr.

The contrast that makes the point is the harmonic approximation, the second-order Taylor expansion about the minimum:

V_harm(r) = k/2 · (r − r_e)², with k = 2·D_e·a²

The harmonic form is exact in the limit of infinitesimal displacement and is what a classical force field's bond term typically uses. It is also qualitatively wrong everywhere else: it is symmetric, it rises without bound, and it therefore cannot dissociate. That failure is the whole reason a fixed functional form is a liability, and the reason MLIPs are built without one.

Grounding and honesty. The Morse functional form itself is textbook and predates this substrate's coverage window, so it is presented here as the standard closed-form diatomic potential rather than attributed to a paper in the collected corpus — an illustrative model, explicitly flagged. What is grounded is its continued use as the working approximation to a diatomic PES parameterized exactly by the quantities named above: substrate work modelling diatomics under external electric fields approximates the potential-energy surface by "a Morse function," predicting response from field-free observables including "the equilibrium bond length, the bond dissociation energy, the polarizability and dipole moment functions, and the vibrational frequency," validated against direct calculations on nine diatomics [52]. The parameter values used in the figure below are illustrative teaching values, not a fit to any particular molecule.

0.5 1.0 1.5 2.0 2.5 3.0 internuclear separation r, Angstrom 0 1 2 3 4 5 6 7 8 potential energy V(r), eV dissociation asymptote V -> De = 4.75 eV well depth De = 4.75 eV re = 0.74 A (equilibrium bond length) harmonic approximation V = (k/2)(r - re)^2, k = 2*De*a^2 rises without bound - cannot dissociate Morse V = De(1 - e^(-a(r - re)))^2 anharmonic, dissociates A diatomic potential-energy curve: anharmonicity is the whole point
Computed diatomic potential-energy curve: the Morse potential V(r) = De*(1 - exp(-a*(r - re)))^2 with illustrative parameters De = 4.75 eV, re = 0.74 Angstrom, a = 1.95 per Angstrom, plotted against its harmonic approximation V = (k/2)*(r - re)^2 with k = 2*De*a^2 (the exact second-order Taylor expansion about the minimum). Marked: the equilibrium bond length re at the curve minimum, the well depth De, and the dissociation asymptote V -> De as r -> infinity. The two curves are indistinguishable near re - the harmonic term is exact to second order - and diverge as soon as the bond stretches: the harmonic form rises without bound and cannot dissociate, while the Morse form flattens to a finite asymptote. This is the one-dimensional picture of why a fixed functional form limits a classical force field, and what a machine-learning interatomic potential generalizes to the full 3N-dimensional surface. The Morse form is the standard closed-form diatomic potential and is used here as an ILLUSTRATIVE teaching model with illustrative parameters; substrate work uses a Morse function for diatomic potential-energy surfaces parameterized by equilibrium bond length and bond dissociation energy [W2111497158].

What generalizes and what does not. Three things carry over from this picture to a real MLIP: energy is a smooth function of geometry, forces are its negative gradient, and the physically important regions are the ones away from equilibrium. Two things do not. First, dimensionality — a diatomic PES is a curve; an N-atom PES lives in 3N−6 internal coordinates, which is why the atomic-energy decomposition and a local cutoff are structural necessities rather than conveniences [3] [11]. Second, the parameterization — a Morse curve has three interpretable parameters, while an MLIP has 10⁴–10⁷ uninterpretable ones, which is exactly the trade that buys flexibility and costs extrapolation guarantees.

Practical threads

Training data and active learning

An MLIP is only as good as the configurations it was shown, and generating them is the dominant cost [11]. Random sampling is wasteful, so the field converged on active learning: run dynamics with the current model, detect configurations where it is untrustworthy, compute quantum reference only for those, retrain, repeat.

The on-the-fly formulation makes forces "either predicted by Bayesian inference or, if necessary, computed by on-the-fly quantum-mechanical calculations and added to a growing ML database, whose completeness is thus never required" — with progressively fewer quantum calls as chemical processes recur [53]. DP-GEN structures the same loop into explicit exploration, labelling and training stages, producing uniformly accurate PES models "with a minimal number of reference data" [54]. The pattern has since been applied to amorphous and liquid oxides [55], to reactive catalysis with Bayesian force fields [56], and sharpened into hyperactive variants that bias exploration toward informative configurations [57] [58] [59].

Uncertainty quantification

Active learning presupposes a usable uncertainty estimate, and it is also what makes an MLIP safe in production — the model needs to say when it has left its training distribution. Gaussian processes give a variance natively [10] [13]; neural networks need it added, typically via committees of independently trained models whose disagreement proxies for error. The literature has developed quantitative uncertainty metrics that demonstrably control error in network-driven chemical discovery [60], uncertainty-aware message-passing architectures [27], and cheap single-model estimators that avoid paying for an ensemble at inference [61].

Extrapolation and transferability limits

The honest framing of the field's central weakness: MLIPs interpolate superbly and extrapolate unpredictably, and unlike a physically motivated functional form they have no correct asymptotic behaviour to fall back on. Locality is the sharpest instance — purely local models cannot see long-range charge transfer, which motivated fourth-generation charge-equilibration schemes [12] — and strictly local architectures make the accuracy/length-scale trade explicit rather than hiding it [32]. Transferability is therefore an empirical property to be measured, which is what the universal-potential and fine-tuning literature is really testing [48] [50] [51] [31].

Integration into MD engines

A potential that cannot be called from a production integrator is a benchmark, not a tool. The path to adoption has been through the established engines: LAMMPS, explicitly designed so that "adding code for new interatomic potentials" is a supported extension point [4] [18]; large-scale parallel and GPU-accelerated MD infrastructure [5] [6]; and universal force engines that decouple the dynamics driver from the energy evaluator [62]. Downstream property calculations — phonons and thermal transport, for instance — then run unchanged on top [63].

Benchmarks, and their discontents

The field's standard metric is force and energy RMSE on held-out configurations [64] [65]. The most useful critical result in the substrate argues this is measuring the wrong thing: the practical use case is producing realistic MD trajectories, and a benchmark suite spanning water, organic molecules, a peptide and materials shows "how the commonly benchmarked force accuracy is not well aligned with relevant simulation metrics," demonstrating when and how state-of-the-art methods fail [66]. Low force error does not guarantee a stable trajectory, a correct radial distribution function, or a physical free-energy landscape. Purpose-built training and evaluation datasets [45] [41] and architecture evaluations across genuinely different domains [31] are the partial answer.

Frontier and open problems

Foundation models for atomistic simulation. The pretrain-then-fine-tune pattern is now established for materials [48] [50] [49] [51] and organic chemistry [43], and the open questions are the familiar ones in a new domain: what the scaling behaviour is, how to evaluate a general model without a general benchmark, and whether a universal potential fine-tuned on a small set beats a bespoke potential trained on the same budget.

Long-timescale dynamics. Making each force evaluation cheap does not solve the timescale problem — MD is still limited by femtosecond timesteps against phenomena that take microseconds and longer. The complementary approaches are enhanced sampling of rare events [67] and learning coarse-grained variables so that the accelerated model and the resolution reduction are trained together [68] [69].

Reactive chemistry. Bond breaking was the historical province of specially constructed reactive force fields [8] [9]. MLIPs handle it naturally in principle, since nothing in the fit presumes a bond topology, but reactive regions are rare in unbiased sampling — which is precisely why active learning coupled to enhanced sampling has become the standard recipe for catalysis and surface chemistry [56] [70] [71] [72].

Coupling to experiment, and to electronic structure. The convincing validations are increasingly experimental observables rather than held-out DFT: infrared and vibrational spectra from ML-driven MD [73] [31], the thermodynamics of liquid and solid water [74], and spectroscopy of aqueous interfaces including quantum nuclear effects [75]. A parallel line learns the electronic structure itself rather than only the energy — transferable electron densities [76], deep-learning DFT Hamiltonians [77] [78], improved exchange-correlation functionals [79], and direct neural solution of the many-electron Schrödinger equation [80]. Taken together these suggest the eventual target is not a faster surrogate for DFT but a learned replacement for parts of the electronic-structure stack.

Dig deeper in lmmol

Methods siblings in this collection:

  • Physics-informed simulation — the neighbouring question of embedding physical law into learned models, where MLIPs are the atomistic-scale instance of the same bargain.
  • Simulation optimization — what to do once each simulation is cheap enough to run many times, which is the regime MLIPs create.

For molecular entities named here, use lmmol's graph index, all proteins, or all diseases rather than guessing an entity URL.

Implementation & visualization hooks

ml: a reference implementation of the two-stage recipe that defines current practice — an equivariant backbone pretrained on a broad dataset, then fine-tuned on a small system-specific set — with an active-learning loop and a committee-disagreement uncertainty estimate driving which configurations get labelled [29] [30] [54] [60] [51].

lmvideo / diagramkit: two diagrams. First, the architecture lineage as a graph — descriptors → kernels → message passing → equivariance → foundation models — with each node annotated by the constraint it removed [3] [10] [24] [29] [48]. Second, the active-learning loop as a cycle: run MD, flag high-uncertainty configurations, label with DFT, retrain, showing where the quantum cost is actually incurred [53] [54].

Key papers

  1. W2166244948: First-principles simulation: ideas, illustrations and the CASTEP code (cited 11,839×)
  2. W2120145199: QUANTUM ESPRESSO: a modular and open-source software project for quantum simulations of materials (cited 29,071×)
  3. W2025444507: Generalized Neural-Network Representation of High-Dimensional Potential-Energy Surfaces (cited 5,041×)
  4. W3201073812: LAMMPS - a flexible simulation tool for particle-based materials modeling at the atomic, meso, and continuum scales (cited 12,018×)
  5. W1031578623: GROMACS: High performance molecular simulations through multi-level parallelism from laptops to supercomputers (cited 27,536×)
  6. W2048621586: Strong scaling of general-purpose molecular dynamics simulations on GPUs (cited 684×)
  7. W3115142961: Molecular Dynamics Simulations in Drug Discovery and Pharmaceutical Development (cited 646×)
  8. W2005944647: A second-generation reactive empirical bond order (REBO) potential energy expression for hydrocarbons (cited 3,750×)
  9. W2333787883: The ReaxFF reactive force-field: development, applications and future directions (cited 2,310×)
  10. W2083415705: Gaussian Approximation Potentials: The Accuracy of Quantum Mechanics, without the Electrons (cited 3,183×)
  11. W1984087004: Representing potential energy surfaces by high-dimensional neural network potentials (cited 410×)
  12. W3086372695: A fourth-generation high-dimensional neural network potential with accurate electrostatics including non-local charge transfer (cited 510×)
  13. W3189164715: Gaussian Process Regression for Materials and Molecules (cited 1,237×)
  14. W3100571530: Comparison of permutationally invariant polynomials, neural networks, and Gaussian approximation potentials in representing water interactions through many-body expansions. (cited 187×)
  15. W2800301423: Machine Learning a General-Purpose Interatomic Potential for Silicon (cited 473×)
  16. W2197007850: Moment Tensor Potentials: A Class of Systematically Improvable Interatomic Potentials (cited 1,515×)
  17. W26088913: Spectral neighbor analysis method for automated generation of quantum-accurate interatomic potentials (cited 1,152×)
  18. W4362665674: FitSNAP: Atomistic machine learning with LAMMPS (cited 48×)
  19. W2910857709: Atomic cluster expansion for accurate and transferable interatomic potentials (cited 947×)
  20. W3176386153: Performant implementation of the atomic cluster expansion (PACE) and application to copper and silicon (cited 250×)
  21. W4387866108: ACEpotentials.jl: A Julia implementation of the atomic cluster expansion (cited 41×)
  22. W4400776655: Cartesian atomic cluster expansion for machine learning interatomic potentials (cited 58×)
  23. W2726670313: SchNet: A continuous-filter convolutional neural network for modeling quantum interactions (cited 470×)
  24. W2778051509: SchNet – A deep learning architecture for molecules and materials (cited 2,245×)
  25. W4300537282: SchNet: A continuous-filter convolutional neural network for modeling\n quantum interactions (cited 344×)
  26. W4287828570: Directional Message Passing for Molecular Graphs (cited 354×)
  27. W4287586570: Fast and Uncertainty-Aware Directional Message Passing for Non-Equilibrium Molecules (cited 164×)
  28. W3189651386: NewtonNet: a Newtonian message passing network for deep learning of interatomic potentials and forces (cited 135×)
  29. W4225405705: E(3)-equivariant graph neural networks for data-efficient and accurate interatomic potentials (cited 1,770×)
  30. W4283021369: MACE: Higher Order Equivariant Message Passing Neural Networks for Fast and Accurate Force Fields (cited 295×)
  31. W4385406529: Evaluation of the MACE force field architecture: From medicinal chemistry to materials science (cited 184×)
  32. W4319162181: Learning local equivariant representations for large-scale atomistic dynamics (cited 659×)
  33. W3126679164: Equivariant message passing for the prediction of tensorial properties and molecular spectra (cited 269×)
  34. W4283641078: Equiformer: Equivariant Graph Attention Transformer for 3D Atomistic Graphs (cited 67×)
  35. W4381713642: EquiformerV2: Improved Equivariant Transformer for Scaling to Higher-Degree Representations (cited 86×)
  36. W4390617556: Enhancing geometric representations for molecules with equivariant vector-scalar interactive message passing (cited 95×)
  37. W4406405186: The design space of E(3)-equivariant atom-centred interatomic potentials (cited 174×)
  38. W4280600366: The Design Space of E(3)-Equivariant Atom-Centered Interatomic Potentials (cited 102×)
  39. W2541404351: ANI-1: an extensible neural network potential with DFT accuracy at force field computational cost (cited 2,054×)
  40. W2800168263: ANI-1, A data set of 20 million calculated off-equilibrium conformations for organic molecules (cited 338×)
  41. W3022256936: The ANI-1ccx and ANI-1x data sets, coupled-cluster and density functional theory properties for molecules (cited 280×)
  42. W2954088480: Approaching coupled cluster accuracy with a general-purpose neural network potential through transfer learning (cited 657×)
  43. W4410498985: MACE-OFF: Short-Range Transferable Machine Learning Force Fields for Organic Molecules (cited 177×)
  44. W4390306067: MACE-OFF: Transferable Short Range Machine Learning Force Fields for Organic Molecules (cited 66×)
  45. W4313545414: SPICE, A Dataset of Drug-like Molecules and Peptides for Training Machine Learning Potentials (cited 179×)
  46. W1992985800: Commentary: The Materials Project: A materials genome approach to accelerating materials innovation (cited 12,959×)
  47. W2949095042: Graph Networks as a Universal Machine Learning Framework for Molecules and Crystals (cited 1,481×)
  48. W4310135808: A universal graph deep learning interatomic potential for the periodic table (cited 902×)
  49. W4282041110: Towards universal neural network potential for material discovery applicable to arbitrary combination of 45 elements (cited 423×)
  50. W4386740800: CHGNet as a pretrained universal neural network potential for charge-informed atomistic modelling (cited 827×)
  51. W4401443596: Data-efficient fine-tuning of foundational models for first-principles quality sublimation enthalpies (cited 49×)
  52. W2111497158: The chemical bond in external electric fields: Energies, geometries, and vibrational Stark shifts of diatomic molecules (cited 159×)
  53. W2057858097: Molecular Dynamics with On-the-Fly Machine Learning of Quantum-Mechanical Forces (cited 652×)
  54. W2898391453: Active learning of uniformly accurate interatomic potentials for materials simulation (cited 502×)
  55. W2981603605: Machine-learned interatomic potentials by active learning: amorphous and liquid hafnium dioxide (cited 183×)
  56. W4294316513: Active learning of reactive Bayesian force fields applied to heterogeneous catalysis dynamics of H/Pt (cited 144×)
  57. W4386714071: Hyperactive learning for data-driven interatomic potentials (cited 103×)
  58. W2916338083: Active learning in materials science with emphasis on adaptive sampling using uncertainties for targeted design (cited 726×)
  59. W2800722845: A strategy to apply machine learning to small datasets in materials science (cited 755×)
  60. W2959623824: A quantitative uncertainty metric controls error in neural network-driven chemical discovery (cited 239×)
  61. W4367174359: Fast uncertainty estimates in deep learning interatomic potentials (cited 72×)
  62. W2886120629: i-PI 2.0: A universal force engine for advanced molecular simulations (cited 384×)
  63. W4310714033: First-principles Phonon Calculations with Phonopy and Phono3py (cited 1,721×)
  64. W2594183968: MoleculeNet: a benchmark for molecular machine learning (cited 3,072×)
  65. W3125542198: Benchmarking graph neural networks for materials chemistry (cited 293×)
  66. W4306313211: Forces are not Enough: Benchmark and Critical Evaluation for Machine Learning Force Fields with Molecular Simulations (cited 167×)
  67. W2167468911: Enhancing Important Fluctuations: Rare Events and Metadynamics from a Conceptual Viewpoint (cited 698×)
  68. W3036877262: Machine learning force fields and coarse-grained variables in molecular\n dynamics: application to materials and biological systems (cited 230×)
  69. W4412521513: Navigating protein landscapes with a machine-learned transferable coarse-grained model (cited 47×)
  70. W4389305212: Machine Learning Interatomic Potentials for Reactive Hydrogen Dynamics at Metal Surfaces Based on Iterative Refinement of Reaction Probabilities (cited 43×)
  71. W4396999021: A reactive neural network framework for water-loaded acidic zeolites (cited 42×)
  72. W4405577576: Data efficient machine learning potentials for modeling catalytic reactivity via active learning and enhanced sampling (cited 50×)
  73. W2614157797: Machine learning molecular dynamics for the simulation of infrared spectra (cited 515×)
  74. W2900475422: Ab initio thermodynamics of liquid and solid water (cited 368×)
  75. W4385062438: First-principles spectroscopy of aqueous interfaces using machine-learned electronic and quantum nuclear effects (cited 65×)
  76. W2889703828: Transferable Machine-Learning Model of the Electron Density (cited 271×)
  77. W4283333204: Deep-learning density functional theory Hamiltonian for efficient ab initio electronic-structure calculation (cited 231×)
  78. W4387405990: Transferable equivariant graph neural networks for the Hamiltonians of molecules and solids (cited 80×)
  79. W3093036756: Quantum chemical accuracy from density functional approximations via machine learning (cited 368×)
  80. W2972153210: <i>Ab initio</i> solution of the many-electron Schrödinger equation with deep neural networks (cited 472×)