Generating Minimal Training Sets for Machine Learned Potentials

TL;DR

This paper introduces a novel method using random network distillation to identify minimal, uncorrelated atomic configurations, significantly reducing the data needed for training accurate machine-learned inter-atomic potentials, demonstrated on molten salts.

Contribution

It presents a new approach combining RND with DFT workflows to efficiently select minimal training data for machine-learned potentials, reducing computational costs.

Findings

Accurately models molten salts with as few as 32 configurations.

Reduces required training data by at least an order of magnitude.

Demonstrates effectiveness on KCl and NaCl systems.

Abstract

This letter presents a novel approach for identifying uncorrelated atomic configurations from extensive data sets with a non-standard neural network workflow known as random network distillation (RND) for training machine-learned inter-atomic potentials (MLPs). This method is coupled with a DFT workflow wherein initial data is generated with cheaper classical methods before only the minimal subset is passed to a more computationally expensive ab initio calculation. This benefits training not only by reducing the number of expensive DFT calculations required but also by providing a pathway to the use of more accurate quantum mechanical calculations for training. The method's efficacy is demonstrated by constructing machine-learned inter-atomic potentials for the molten salts KCl and NaCl. Our RND method allows accurate models to be fit on minimal data sets, as small as 32 configurations, reducing the required structures by at least one order of magnitude compared to alternative methods.