Papers for

materials simulation engineers

Papers whose findings have a practical use for this group, as judged from the abstract. Open a paper to read what it means in practice.

Multi cutoff method improves prediction accuracy in atom modeling

MCSA: Multi-Cutoff Spectral Aggregation in SO(3) Equivariant GNNs for Interatomic Potentials

Abstract: Machine Learning Interatomic Potentials (MLIPs) based on Equivariant Graph Neural Networks (EGNNs) have achieved remarkable progress in atomistic modeling. However, accurately representing atomic interactions across different spatial scales remains challenging. In this work, we investigate the radial spectral properties of SO(3)-equivariant message passing and establish that the cutoff radius controls the characteristic spatial-frequency scale of radial interactions. Based on this analysis, we propose Multi-Cutoff Spectral Aggregation (MCSA), a plug-and-play module that aggregates radial kernels with different cutoff radii to enrich multi-scale geometric representations. To achieve efficient implementation, we exploit the compact support of radial kernels to fuse cutoff-dependent radial weights before angular tensor-product operations, avoiding redundant angular computations while preserving rotational equivariance. Extensive experiments across representative EGNN backbones demonstrate consistent improvements in energy, force, and stress prediction, as well as enhanced accuracy in higher-order properties, including phonon dispersion and lattice thermal conductivity. Further evaluations on layered materials with distinct interaction scales demonstrate the effectiveness of physically motivated cutoff selection. Code is available at: https://github.com/JF-Tan/MCSA.

Sun 8 NovMachine LearningArtificial Intelligence
The gist
Predicting how atoms interact is important for designing new materials, but it can be tricky because atoms influence each other at different distances. The authors found that changing a key setting, called the cutoff radius, affects how well the model captures these interactions. They created a way to combine information from multiple cutoff distances, improving the model’s ability to understand interactions at different scales. This leads to better predictions of energy, forces, and other material properties.
Open → 2611.11226v1

Benchmarks reveal scaffold effects and execution checks in autonomous materials simulation

DFT-Agent Bench: Scaffold Effects and Execution Provenance in Autonomous DFT

Abstract: Large language models are increasingly deployed as autonomous agents for first-principles materials simulation. We report empirical findings from DFT-Agent Bench, a benchmark of 200+ Quantum ESPRESSO tasks spanning relaxation, electronic structure, equation-of-state, and elastic-tensor pipelines over multiple material tiers. Each task carries a QE-derived ground truth with public-reference cross-checks where available, and a provenance audit checks whether the required QE execution contract was met. Under one execution harness, we compare a near-bare ReAct baseline with a bundled v5.1 scaffold on two models. The complete scaffold improves bands+DOS by +0.08/+0.11 and d-electron vc-relax by +0.17 on both models, while high-scoring relaxation and equation-of-state cells leave little room for improvement. The elastic result is model-dependent (0.18 to 0.40 for deepseek-v4-pro and 0.62 to 0.59 for claude-opus-4-8). A fixed 100-task analysis with retrieval disabled supports the aggregate pattern: the fully frozen DeepSeek comparison gains +0.17, while the Opus comparison gains +0.10 against a historical pool baseline (p < 0.001). The provenance gate catches incomplete or unmarked executions that can receive high rubric scores; it is an execution-contract check, not a test for intentional fraud. These results characterize the evaluated DFT workflows. The anonymous review artifact contains the frozen diagnostic slice, prompts, scorer, and solver revisions: https://anonymous.4open.science/r/231e31re91287h/README.md.

Sat 7 NovArtificial Intelligence
The gist
Simulating materials at the atomic level helps scientists understand their properties, but these simulations are complex and require careful computer instructions. The authors tested different ways that AI language models guide these simulations, comparing a simple baseline with a more detailed approach called a scaffold. They found the scaffold improved accuracy for many tasks and also developed a way to check that the computer actually ran the full simulation. Their work helps show how AI can better assist in these materials simulations.
Open → 2611.10629v1

High performance molecular dynamics on Nvidia and Intel GPUs with versatile atomic models

Isothermal-Isobaric Molecular Dynamics with a Foundation Interatomic Potential in LAMMPS across NVIDIA and Intel GPUs

Abstract: Foundation machine-learning interatomic potentials (MLIPs) offer one model for many chemistries, but their deep, globally coupled architectures do not fit the spatial decomposition that scales classical and short-ranged MLIPs. We present pair_style uma, a LAMMPS pair style that evaluates the Universal Model for Atoms (UMA) through LibTorch on NVIDIA and Intel GPUs from a single source tree. It reimplements FairChem's edge-sharded graph parallelism in C++ over exported TorchScript modules, with one MPI rank per device, NCCL or oneCCL collectives, and the whole energy and force path in double precision. The strain-gradient virial is carried through the exported modules and the cross-device reductions, so that fix npt runs on both vendors' hardware. We measure NPT capacity and cost on one 4xA100-40GB node, one 8xH200 node, and one 12-tile Intel node, and on up to 32 Intel nodes. Without activation checkpointing, one node holds NaCl cells of about 8,000 atoms (A100, 0.21 ns/day), 22,000 atoms (Intel, 0.10 ns/day), and 47,000 atoms (H200, 0.16 ns/day). Graph parallelism raises the capacity of a single device by 8.0x on H200 and 4.6x on A100. Across nodes, capacity grows to about 779,000 atoms on 32 Intel nodes, with a slower speed at 0.007 ns/day. On H200 and A100, a FLOP count of FairChem's graph-parallel predictor gives lower bounds of 31% and 51% of FP64 peak. From one omat checkpoint and with no refit, nine 100 ps NPT heating ramps (NaCl, equiatomic Pt-Ru, and equiatomic (HfNbTaTiZr)N, on all three node types) ran to completion.

Mon 2 NovComputational Engineering, Finance, and Science
The gist
Running realistic simulations of materials requires complex models that predict atomic interactions, but these models can be hard to run efficiently on GPUs. The authors developed a specialized software tool that allows a broad class of machine-learned atomic models to run on both Nvidia and Intel GPUs using the popular LAMMPS simulation code. They demonstrate that their implementation scales to large simulations involving hundreds of thousands of atoms and runs at practical speeds across different hardware types. This approach makes it easier and faster to perform advanced molecular dynamics simulations with a unified software setup.
Open → 2611.03892v1

Machine learning improves prediction of molecular electronic properties

Where the Error Lives: Spectral Representations of Learned Kohn-Sham Hamiltonians

Abstract: Kohn-Sham density functional theory (DFT) is the standard method for computing the electronic structure of molecules and materials, but its self-consistent-field (SCF) iterations make it expensive for large systems and high-throughput screening. Machine-learned Hamiltonian models offer a shortcut: an equivariant network predicts the Kohn-Sham Hamiltonian matrix in one forward pass, and diagonalizing the prediction yields orbital energies, the HOMO-LUMO gap, molecular orbitals and the electron density. These models are trained to match the reference matrix entry by entry, yet a small entry-wise error does not guarantee accurate eigenvalues or eigenvectors. Two physics-informed alternatives go partway: a recently proposed wavefunction-alignment loss adds a spectral penalty without diagonalizing the prediction, and matching the electron density of the diagonalized prediction supervises it directly. Neither is complete: by representing the prediction error in the eigenbasis of the reference Hamiltonian, we show that each sees only part of the error, and that shifting the predicted orbital energies leaves the electron density exactly unchanged. The same analysis identifies a complete training signal: the cumulative local density of states (cumulative LDOS), the number of electrons at each point in space with energy below a given threshold. The cumulative LDOS is a single physical observable that contains the electron density as a special case, responds to every component of the error, determines the Hamiltonian uniquely, and yields gradients that do not vanish for large orbitalenergy errors. Fine-tuning three Hamiltonian networks on QH9 in this way, which we call FermiScope, reduces the HOMO-LUMO gap error without increasing the element-wise error. The gains carry over to out-of-distribution and molecular-dynamics splits, where the wavefunction-alignment loss helps less.

Tue 27 OctArtificial Intelligence
The gist
Calculating how electrons behave in molecules is important but usually takes a lot of computer time. The authors studied how machine-learning models can predict these behaviors faster by guessing part of the math involved. They found that using a new way to measure errors, based on how many electrons have energies below certain levels, helps the model make better predictions about important properties like the gap between electron energy levels. This approach works better than older methods and can apply to new molecules the model has not seen before.
Open → 2610.37843v1

Limits of machine learning models for atomic simulations in high dimensions

On the Theoretical Limits of Learning Unconstrained Interatomic Potentials

Abstract: Machine-learned interatomic potentials (MLIPs) accelerate molecular dynamics (MD) simulations by learning the energies and forces that govern atomic motion. Recent work has shown that relaxing physical inductive biases can improve computational efficiency and predictive accuracy, raising the question of whether energy-and-force matching alone is sufficient to ensure accurate sampling. We show that, for unconstrained hypothesis classes, the relationship between prediction error and sampling accuracy is exponentially ill-conditioned in dimension. In particular, prediction errors that decay exponentially with dimension can coexist with nearly maximal total variation distance between the learned and target distributions. We identify two failure mechanisms. The first is kinetic trapping, where relaxation to an accurate learned equilibrium requires exponentially long times, even when simulations are initialized from training configurations at which energies and forces are matched exactly. The second is equilibrium mismatch, where the learned stationary distribution itself remains far from the target. Under explicit assumptions, we show that equilibrium mismatch can persist even in local, symmetry-preserving models, where small per-atom errors coexist with a distributional discrepancy that approaches its maximum as system size increases. Finally, we establish global energy-control conditions that guarantee equilibrium accuracy and identify additional conditions required for rapid relaxation.

Sat 10 OctMachine Learning
The gist
Machine-learned models are used to speed up simulations of how atoms move by predicting their energies and forces. This paper shows that when these models have no specific physical constraints, their predictions may look accurate but can still lead to very wrong behaviors in actual simulations, especially as the number of atoms increases. Two key problems arise: simulations get stuck for a very long time, or they settle into incorrect stable states. The authors also provide conditions that, if met, guarantee accurate and fast simulations.
Open → 2610.15078v1

Neural network predicts molecular energy across multiple shapes

NucleoNet: A Nuclear-Aware Determinant Ansatz for Multi-Geometry Neural Variational Monte Carlo

Abstract: Neural-network variational Monte Carlo provides an accurate route to solving the electronic Schrödinger equation, but many neural wavefunctions are optimized for a single nuclear geometry, requiring separate calculations to map out a potential energy surface. We introduce NucleoNet, a nuclear-aware determinant ansatz in which electrons and nuclei form a unified particle set propagated through a shared electron--nucleus backbone that jointly evolves one-particle and two-particle representations. Nuclei carry learned representations that evolve alongside the electronic representations and condition orbital construction, while antisymmetry is imposed only by electronic determinants. On equilibrium small-molecule benchmarks, NucleoNet achieves accuracy comparable to established neural-wavefunction baselines with substantially fewer parameters. For each molecular system, a single model optimized jointly across geometries describes the N$_2$ dissociation curve, Li displacement in LiCN, and the ethane and formamide minimum-energy paths for bond dissociation. Sparse-geometry training generalizes to unseen geometries for N$_2$ and LiCN without further optimization, and N$_2$ ablations show that nuclear streams substantially improve this interpolation. These results show that a compact wavefunction with jointly evolved nuclear representations can describe the geometries of a molecular system with a single set of parameters.

Fri 9 OctMachine Learning
The gist
Calculating how molecules behave when their shapes change is important for understanding chemistry, but usually requires repeating complex calculations for each shape. The authors present NucleoNet, a method that trains a single neural network to handle many molecule shapes at once by treating electrons and nuclei together in a shared model. This approach simplifies predictions and can smoothly estimate energies for new shapes the model hasn’t seen before. It works well on small molecules, using fewer parameters than other methods with similar accuracy.
Open → 2610.12985v1

Neural model predicts ionic structures in polymer electrolytes accurately

A structure-preserving neural density functional for the ions of a polymer electrolyte

Abstract: Predicting the structure and response of inhomogeneous polymer electrolytes requires a description of ion correlations that retains molecular-scale accuracy while remaining transferable across spatial scales and geometries. We develop a neural density functional for electrolytes that preserves spatial symmetries, thermodynamic integrability and the Noether identities, with perfect screening recovered in stable, noncritical bulk states. Its nonlinear density dependence captures the concentration-dependent correlations missed by a pair closure, including a crossover from enhanced to suppressed long-wavelength number fluctuations at strong coupling. The functional describes density profiles at an untrained salt concentration and predicts bulk structure factors and the long-wavelength number response. Trained solely on planar density and internal-force profiles from molecular dynamics, the functional predicts ionic structure in larger domains and in two-dimensional external fields. On the same ion data, it is more accurate than three other neural density-functional architectures and keeps its accuracy with a quarter of the training runs, where the errors of the best alternative grow by about two thirds. The spatial transferability provides a necessary foundation for connecting molecular correlations to continuum predictions at larger scales.

Thu 8 OctMachine Learning
The gist
It is hard to predict how ions behave inside polymer electrolytes, which are materials used in batteries and other devices. The authors create a neural network model that respects the physical laws and symmetries of the system, learning from simulations of ions near flat surfaces. This model can then predict ion arrangements in more complex shapes and different conditions better than other methods. It works well even when trained with less data and helps link small-scale ion behavior to larger-scale material properties.
Open → 2610.12132v1

Crystallization in hard-core lattice models occurs at low and high activity

A general framework for crystallization in maximal hard-core models

Abstract: We study general maximal hard-core models on periodic lattice-type graphs. While for standard models, phase coexistence may arise for high activity, for maximal models this behavior may occur also for sufficiently low activity values. Relying on the concept of volume allocation, we develop a {\em unified set of assumptions} that imply the Peierls condition, a convergent cluster expansion for the partition function, and hence the conclusions of Pirogov-Sinai theory for this class of models. We further check these assumptions for a number of examples including standard lattice-type periodic graphs, proving crystallization for both high and low activity. We also derive estimates for the values of activity that bound the phase coexistence regions in the phase diagram, rewriting for this purpose the proof of a technical result in the derivation of Pirogov-Sinai theory.

Thu 8 OctInformation Theory
The gist
This paper studies how particles arranged on a repeating grid cannot be too close together, a rule called the hard-core model. The researchers look at a version called maximal hard-core models, where particles fill as much space as possible without touching. They show that these systems can form stable patterns, called crystals, not only when particles are very active but also when activity is low. The authors develop a general mathematical framework that explains this behavior and apply it to many examples, clarifying when these patterns appear.
Open → 2610.11042v1

Physics aligned models improve predictions of electronic ground states

Physics-Aligned Electronic Ground-State Learning Improves Generalization

Abstract: Machine-learned interatomic potentials (MLIPs) excel at in-distribution tasks, accelerating drug and material development, yet they struggle to generalize out-of-distribution. We propose to push the cost-accuracy Pareto frontier by designing observable-agnostic electronic ground-state descriptor models (GSMs) with computational costs situated between MLIPs and Kohn-Sham density functional theory (KS-DFT). We align the learning objectives and architectures of GSMs with the governing equations of KS-DFT by enforcing physical constraints and removing optimization pressure on unphysical or irrelevant degrees of freedom. In our size-extrapolation experiments from QM9 to QM40, our combined contributions OrthoNormal-Loss (ON-Loss) and Grassmann Restricted Occupied-Orbital Training (GROOT) reach a 79.1% energy and 83.4% force mean absolute error (MAE) reduction over previous state-of-the-art density GSMs. For Hamiltonian GSMs, ON-Loss and Residual Optimal-gauge Conditioning-aware KS-Eq. Training (ROCKET) together reduce the energy and force MAEs of the strongest baseline by 99.8% and 95.9%, respectively. Using a self-consistency rejection criterion, we filter out extrapolation errors on QMugs, rejecting fewer than 0.4% of predictions while reaching an energy MAE of 0.07 mHa. Finally, we demonstrate the efficiency of label-free self-consistency fine-tuning, and transfer GSMs to reactive chemistry in Transition1x, reaching energy errors below chemical accuracy.

Wed 7 OctMachine Learning
The gist
Predicting the behavior of molecules and materials accurately and quickly is hard because standard methods either take too long or don’t handle new, unseen molecules well. The authors propose new machine learning models that respect the physics of how electrons arrange themselves in molecules, improving prediction accuracy and speed. These models better generalize to larger or different molecules, and can self-check their own predictions to avoid big mistakes. They also show the models can adapt to chemical reactions without needing additional labeled data.
Open → 2610.10298v1

FlashCart speeds up accurate atom simulations with efficient tensor products

FlashCart: Fast Cartesian Tensor Products for Equivariant Interatomic Potentials

Abstract: Machine-learned interatomic potentials extend atomistic simulations beyond the length- and timescales accessible to electronic-structure methods. However, the computational cost of equivariant architectures limits the local correlations they can represent in practice and therefore their achievable accuracy. Here we introduce FlashCart, which makes higher-order correlations affordable by combining generated GPU kernels with an architecture that recursively builds equivariant features and compresses them to a fixed width at each step. We express tensor products in independent Cartesian components and symbolically simplify them and their derivatives, producing fused kernels that often outperform optimized spherical counterparts. We then show that increasing correlation order improves accuracy more efficiently than increasing width, depth, or tensor rank. On SPICE-MACE-OFF, FlashCart models advance the measured accuracy-efficiency frontier: a model with $5.6$ million parameters achieves lower energy and force errors and $10\times$ faster inference than a transformer with $189$ million parameters.

Mon 5 OctMachine Learning
The gist
Simulating how atoms and molecules interact is important but usually very slow and costly. The authors introduce FlashCart, a technique that makes these simulations faster and more accurate by cleverly organizing and simplifying complex calculations with GPUs. Their approach lets computers understand detailed atomic behaviors without needing huge computational resources, outperforming previous methods with fewer parameters and faster speeds.
Open → 2610.06409v1

Memory restores hidden states for better atom-by-atom crystal evolution

AtomWorld-Mem: Memory-Restored World States for Long-Horizon Atomistic Evolution

Abstract: High-fidelity atomistic evolution over long timescales requires more than observing the current crystal configuration. Instantaneous atomistic snapshots are often incomplete: locally similar configurations can correspond to different hidden dynamical contexts, future event preferences, and waiting-time scales. We argue that this snapshot ambiguity makes long-horizon atomistic evolution fundamentally a memory-based world-state restoration problem. To address this, we introduce AtomWorld-Mem, a memory-restored atomistic world model that recovers the latent world state missing from instantaneous crystal snapshots. AtomWorld-Mem treats the evolving alloy as an AtomWorld: spatial encoders write multi-scale atomistic keyframes from dense local topology and sparse long-range defect context, while short-term event memory and long-term structural memory integrate these keyframes across time to restore a future-predictive evolutionary state. The restored state is used to prioritize legal vacancy-mediated events under single-event Kinetic Monte Carlo (KMC) constraints, while event legality, physical execution, and residence-time updates remain governed by the underlying simulator. Empirically, AtomWorld-Mem improves long-horizon atomistic progress under fixed microscopic event budgets while maintaining high-fidelity evolution across energetic, structural, and vacancy-transport observables. It further transfers zero-shot across diverse unseen alloy-temperature AtomWorlds, suggesting that the learned memory-restoration mechanism captures reusable principles of hidden-state inference rather than a system-specific local energy heuristic. These results position memory-restored world-state modeling as a promising route toward efficient, physically grounded, and transferable atomistic evolution.

Fri 25 SeptArtificial Intelligence
The gist
Predicting how materials like crystals change over time is hard because looking at just one snapshot of atoms doesn’t show the full story. The authors found that keeping memory of past atom arrangements helps reveal hidden information that influences future changes. They built a system called AtomWorld-Mem that remembers previous states to improve long-term predictions of how atoms move and rearrange. This system better matches real material behavior and works well even on materials it wasn’t trained on.
Open → 2609.31133v1

Complete neural initialization speeds up materials density calculations

Complete Neural Electronic Initialization Accelerates Materials DFT

Abstract: We present the first complete machine learning method for accelerating plane-wave density functional theory (DFT) in materials under the projector augmented wave (PAW) formalism. We formalize seven criteria that a \textit{Complete Neural Electronic Initializer} must satisfy for practical end-to-end PAW DFT acceleration. Applying these criteria to prior work reveals two missing structure-dependent components, augmentation occupancies and spin initialization, that prevent existing methods from providing complete reference-free initialization. Controlled ablations show that omitting these components can eliminate or reverse the acceleration obtained via models that only predict the smooth valence density. We satisfy these missing requirements by introducing AugNet, the first general equivariant model for PAW augmentation occupancies, and the first general spin density model for materials, which predicts the smooth spin-difference density and spin-difference PAW augmentation occupancies using predicted magnetic moments to constrain the global magnetic state. Combined with existing valence density models, these components satisfy all seven criteria and form a fully reference-free electronic initializer for materials DFT, requiring no electronic quantities from a converged target calculation. Our method reduces end-to-end DFT wall time by up to ~25% on unseen structures while preserving converged energies.

Fri 18 SeptMachine Learning
The gist
Calculating the properties of materials at the atomic level is very slow because it needs to figure out how electrons behave. The authors found that previous AI methods for speeding this up missed important details. They created new models that include these missing parts and can start the calculations from scratch without extra information. This new approach speeds up the calculations by about a quarter while keeping results accurate.
Open → 2609.21759v1