Robustness of Local Predictions in Atomistic Machine Learning Models

TL;DR

This paper introduces a metric called local prediction rigidity (LPR) to evaluate the robustness of atomistic machine learning models' local predictions, and explores how training data influences this robustness across various chemical systems.

Contribution

It proposes the LPR metric for assessing local prediction robustness and offers strategies to improve model reliability and interpretability.

Findings

LPR effectively measures prediction robustness.

Training data composition impacts LPR.

Strategies can enhance model robustness.

Abstract

Machine learning (ML) models for molecules and materials commonly rely on a decomposition of the global target quantity into local, atom-centered contributions. This approach is convenient from a computational perspective, enabling large-scale ML-driven simulations with a linear-scaling cost, and also allow for the identification and post-hoc interpretation of contributions from individual chemical environments and motifs to complicated macroscopic properties. However, even though there exist practical justifications for these decompositions, only the global quantity is rigorously defined, and thus it is unclear to what extent the atomistic terms predicted by the model can be trusted. Here, we introduce a quantitative metric, which we call the local prediction rigidity (LPR), that allows one to assess how robust the locally decomposed predictions of ML models are. We investigate the dependence of LPR on the aspects of model training, particularly the composition of training dataset, for a range of different problems from simple toy models to real chemical systems. We present strategies to systematically enhance the LPR, which can be used to improve the robustness, interpretability, and transferability of atomistic ML models.