Transfer learning for chemically accurate interatomic neural network potentials

TL;DR

This paper demonstrates that transfer learning, especially discriminative fine-tuning, significantly enhances the efficiency and accuracy of neural network interatomic potentials for organic molecules, enabling large-scale simulations.

Contribution

It introduces a transfer learning approach with discriminative fine-tuning for interatomic potentials, improving sample efficiency and accuracy using pre-training on DFT data.

Findings

Pre-training on DFT data improves sample efficiency.

Fine-tuning with energy labels alone yields accurate forces.

Pre-trained GM-NN potentials can be easily adapted to organic molecules.

Abstract

Developing machine learning-based interatomic potentials from ab-initio electronic structure methods remains a challenging task for computational chemistry and materials science. This work studies the capability of transfer learning, in particular discriminative fine-tuning, for efficiently generating chemically accurate interatomic neural network potentials on organic molecules from the MD17 and ANI data sets. We show that pre-training the network parameters on data obtained from density functional calculations considerably improves the sample efficiency of models trained on more accurate ab-initio data. Additionally, we show that fine-tuning with energy labels alone can suffice to obtain accurate atomic forces and run large-scale atomistic simulations, provided a well-designed fine-tuning data set. We also investigate possible limitations of transfer learning, especially regarding the design and size of the pre-training and fine-tuning data sets. Finally, we provide GM-NN potentials pre-trained and fine-tuned on the ANI-1x and ANI-1ccx data sets, which can easily be fine-tuned on and applied to organic molecules.