Interpolation and differentiation of alchemical degrees of freedom in machine learning interatomic potentials

arXiv:2404.10746v1 Announce Type: cross
Abstract: Machine learning interatomic potentials (MLIPs) have become a workhorse of modern atomistic simulations, and recently published universal MLIPs, pre-trained on large datasets, have demonstrated remarkable accuracy and generalizability. However, the computational cost of MLIPs limits their applicability to chemically disordered systems requiring large simulation cells or to sample-intensive statistical methods. Here, we report the use of continuous and differentiable alchemical degrees of freedom in atomistic materials simulations, exploiting the fact that graph neural network MLIPs …

abstract accuracy arxiv become computational cond-mat.mtrl-sci cost cs.lg datasets differentiation freedom however interpolation large datasets machine machine learning modern physics.chem-ph simulations systems type universal