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.

Active learning of linearly parametrized interatomic potentials

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

Active delta-learning for fast construction of interatomic potentials and stable molecular dynamics simulations

Accelerating the Training and Improving the Reliability of Machine-Learned Interatomic Potentials for Strongly Anharmonic Materials through Active Learning

Machine Learning Inter-Atomic Potentials Generation Driven by Active Learning: A Case Study for Amorphous and Liquid Hafnium dioxide

A transferable active-learning strategy for reactive molecular force fields

Committee neural network potentials control generalization errors and enable active learning

Utilizing Machine Learning for Efficient Parameterization of Coarse Grained Molecular Force Fields

Active learning strategies for atomic cluster expansion models

Physics-informed active learning for accelerating quantum chemical simulations

On-the-fly active learning of interpretable Bayesian force fields for atomistic rare events

The MLIP package: Moment Tensor Potentials with MPI and Active Learning

Uncertainty-biased molecular dynamics for learning uniformly accurate interatomic potentials

Machine-learned interatomic potentials by active learning: amorphous and liquid hafnium dioxide

Learning Interatomic Potentials at Multiple Scales

Accelerating high-throughput searches for new alloys with active learning of interatomic potentials

Optimal design of experiments in the context of machine-learning inter-atomic potentials: improving the efficiency and transferability of kernel based methods

De novo exploration and self-guided learning of potential-energy surfaces

MLIP-3: Active learning on atomic environments with Moment Tensor Potentials