no code implementations • 1 Feb 2024 • Joshua A. Vita, Amit Samanta, Fei Zhou, Vincenzo Lordi
Model ensembles are effective tools for estimating prediction uncertainty in deep learning atomistic force fields.
no code implementations • 4 Oct 2023 • Joshua A. Vita, Dallas R. Trinkle
While machine learning (ML) interatomic potentials (IPs) are able to achieve accuracies nearing the level of noise inherent in the first-principles data to which they are trained, it remains to be shown if their increased complexities are strictly necessary for constructing high-quality IPs.
no code implementations • 12 Feb 2023 • Joshua A. Vita, Daniel Schwalbe-Koda
In this work, we show how architectural and optimization choices influence the generalization of NNIPs, revealing trends in molecular dynamics (MD) stability, data efficiency, and loss landscapes.