Saved in:
Bibliographic Details
Main Authors: Novelli, Pietro, Meanti, Giacomo, Buigues, Pedro J., Rosasco, Lorenzo, Parrinello, Michele, Pontil, Massimiliano, Bonati, Luigi
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2505.05652
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912367222194176
author Novelli, Pietro
Meanti, Giacomo
Buigues, Pedro J.
Rosasco, Lorenzo
Parrinello, Michele
Pontil, Massimiliano
Bonati, Luigi
author_facet Novelli, Pietro
Meanti, Giacomo
Buigues, Pedro J.
Rosasco, Lorenzo
Parrinello, Michele
Pontil, Massimiliano
Bonati, Luigi
contents Training machine learning interatomic potentials that are both computationally and data-efficient is a key challenge for enabling their routine use in atomistic simulations. To this effect, we introduce franken, a scalable and lightweight transfer learning framework that extracts atomic descriptors from pretrained graph neural networks and transfer them to new systems using random Fourier features-an efficient and scalable approximation of kernel methods. Franken enables fast and accurate adaptation of general-purpose potentials to new systems or levels of quantum mechanical theory without requiring hyperparameter tuning or architectural modifications. On a benchmark dataset of 27 transition metals, franken outperforms optimized kernel-based methods in both training time and accuracy, reducing model training from tens of hours to minutes on a single GPU. We further demonstrate the framework's strong data-efficiency by training stable and accurate potentials for bulk water and the Pt(111)/water interface using just tens of training structures. Our open-source implementation (https://franken.readthedocs.io) offers a fast and practical solution for training potentials and deploying them for molecular dynamics simulations across diverse systems.
format Preprint
id arxiv_https___arxiv_org_abs_2505_05652
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Fast and Fourier Features for Transfer Learning of Interatomic Potentials
Novelli, Pietro
Meanti, Giacomo
Buigues, Pedro J.
Rosasco, Lorenzo
Parrinello, Michele
Pontil, Massimiliano
Bonati, Luigi
Computational Physics
Training machine learning interatomic potentials that are both computationally and data-efficient is a key challenge for enabling their routine use in atomistic simulations. To this effect, we introduce franken, a scalable and lightweight transfer learning framework that extracts atomic descriptors from pretrained graph neural networks and transfer them to new systems using random Fourier features-an efficient and scalable approximation of kernel methods. Franken enables fast and accurate adaptation of general-purpose potentials to new systems or levels of quantum mechanical theory without requiring hyperparameter tuning or architectural modifications. On a benchmark dataset of 27 transition metals, franken outperforms optimized kernel-based methods in both training time and accuracy, reducing model training from tens of hours to minutes on a single GPU. We further demonstrate the framework's strong data-efficiency by training stable and accurate potentials for bulk water and the Pt(111)/water interface using just tens of training structures. Our open-source implementation (https://franken.readthedocs.io) offers a fast and practical solution for training potentials and deploying them for molecular dynamics simulations across diverse systems.
title Fast and Fourier Features for Transfer Learning of Interatomic Potentials
topic Computational Physics
url https://arxiv.org/abs/2505.05652