In operando active learning of interatomic interaction during large-scale simulations

Abstract

A well-known drawback of state-of-the-art machine-learning interatomic potentials is their poor ability to extrapolate beyond the training domain. For small-scale problems with tens to hundreds of atoms this can be solved by using active learning which is able to select atomic configurations on which a potential attempts extrapolation and add them to the ab initio-computed training set. In this sense an active learning algorithm can be viewed as an on-the-fly interpolation of an ab initio model. For large-scale problems, possibly involving tens of thousands of atoms, this is not feasible because one cannot afford even a single density functional theory computation with such a large number of atoms. This work marks a new milestone toward fully automatic ab initio-accurate large-scale atomistic simulations. We develop an active learning algorithm that identifies local subregions of the simulation region where the potential extrapolates. Then the algorithm constructs periodic configurations out of these local, non-periodic subregions, sufficiently small to be computable with plane-wave density functional theory codes, in order to obtain accurate ab initio energies. We benchmark our algorithm on the problem of the screw dislocation motion in bcc tungsten and show that our algorithm reaches ab initio accuracy, down to typical magnitudes of numerical noise in DFT codes. We show that our algorithm reproduces material properties such as core structure, Peierls barrier, and Peierls stress. This unleashes new capabilities for computational materials science toward applications which have currently been out of scope if approached solely by ab initio methods.

Published version

References in corpus (19)

• Gaussian Approximation Potentials: the accuracy of quantum mechanics, without the electrons
• On representing chemical environments
• ANI-1: An extensible neural network potential with DFT accuracy at force field computational cost
• Moment Tensor Potentials: a class of systematically improvable interatomic potentials
• A Spectral Analysis Method for Automated Generation of Quantum-Accurate Interatomic Potentials
• Less is more: sampling chemical space with active learning
• Active learning of linearly parametrized interatomic potentials
• On-the-fly machine learning force field generation: Application to melting points
• Active Learning of Uniformly Accurate Inter-atomic Potentials for Materials Simulation
• Accelerating crystal structure prediction by machine-learning interatomic potentials with active learning
• FCHL revisited: faster and more accurate quantum machine learning
• Achieving DFT accuracy with a machine-learning interatomic potential: thermomechanics and defects in bcc ferromagnetic iron
• A learning scheme to predict atomic forces and accelerate materials simulations
• Machine-learned multi-system surrogate models for materials prediction
• Machine learning of molecular properties: locality and active learning
• Multiscale simulations in simple metals: a density-functional based methodology
• The influence of transition metal solutes on dislocation core structure and values of Peierls stress and barrier in tungsten
• Computing energy barriers for rare events from hybrid quantum/classical simulations through the virtual work principle
• Geometries of edge and mixed dislocations in bcc Fe from first principles calculations

Cited by in corpus (8)

• Efficiency, Accuracy, and Transferability of Machine Learning Potentials: Application to Dislocations and Cracks in Iron
• Machine-learning potentials for nanoscale simulations of deformation and fracture: example of TiB${}_2$ ceramic
• Machine-learning potentials enable predictive $\text{and}$ tractable high-throughput screening of random alloys
• Modeling Extensive Defects in Metals through Classical Potential-Guided Sampling and Automated Configuration Reconstruction
• Unsupervised learning-based structural analysis: Search for a characteristic low-dimensional space by local structures in atomistic simulations
• Ab initio framework for deciphering trade-off relationships in multi-component alloys
• Automated atomistic simulations of dissociated dislocations with ab initio accuracy
• Crystal Nucleation in Eutectic Al-Si Alloys by Machine-Learned Molecular Dynamics