Learning Between the Lines: Interaction Interpolation in a Multi Element Machine Learning Interatomic Potential
Abstract
The continuous development of machine learning inter-atomic potentials (MLIAPs)has greatly accelerated materials research in recent years. However, the need to con- struct custom MLIAPs for each unique problem remains a bottleneck which raises the question of the transferability of MLIAPs. While foundation MLIAPs can achieve transferability due to the broad coverage of chemical space represented in their training sets, their size and computational cost limits their routine use. We address this gap by developing a lightweight general-purpose MLIAP for metal alloys spanning twelve elements (Al, Sc, Ti, V, Cu, Zn, Zr, Nb, Mo, Hf, Ta, W). Here, we utilize a step-wise active learning (AL) approach to develop a compact yet diverse dataset and model that generalizes well by interpolating from previously seen interactions. The resulting potentials transfer effectively to unseen, higher body-order alloys systems, requiring only unary and binary training data to capture ternary and quaternary systems. The model also transfers well to new elements via an alchemical extrapolation within the network, requiring only unary data for the new element and its alloys. These findings suggest that MLIAPs trained on targeted, diversity-driven datasets provide transferability at a relatively low computational cost, offering a practical path toward general-purpose, resource-efficient MLIAPs for materials modeling.
Related
- PASS: Perturbation augmented space group structure sampling for transferable Fe-O machine learning interatomic potential
- Precipitate phase selection and grain boundary morphology in Cu-Ni-Si-Mn alloys: A machine-learning interatomic potential study
- Machine-Learned Interatomic Potential for the Mo-S-Al-O System
- A GRACE Interatomic Potential for W--Ta Systems: Application to Ta Segregation and Pinning of Shear-Coupled Grain-Boundary Migration
- Agent-E2MD: Autonomous Translation of Interatomic Potential Equations into Physically Validated Pair Styles for Molecular Dynamics in LAMMPS
BibTeX
@misc{soyemi2026learning,
title = {{Learning Between the Lines: Interaction Interpolation in a Multi Element Machine Learning Interatomic Potential}},
author = {Ademola Soyemi and Kipton Barros and Timothy C. Germann and Benjamin T. Nebgen and Brenden W. Hamilton},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.38876},
eprint = {2610.38876},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.38876},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Learning Between the Lines: Interaction Interpolation in a Multi Element Machine Learning Interatomic Potential
AU - Ademola Soyemi
AU - Kipton Barros
AU - Timothy C. Germann
AU - Benjamin T. Nebgen
AU - Brenden W. Hamilton
PY - 2026
DA - 2026/10/28
PB - arXiv
DO - 10.48550/arxiv.2610.38876
UR - https://arxiv.org/abs/2610.38876
AB - The continuous development of machine learning inter-atomic potentials (MLIAPs)has greatly accelerated materials research in recent years. However, the need to con- struct custom MLIAPs for each unique problem remains a bottleneck which raises the question of the transferability of MLIAPs. While foundation MLIAPs can achieve transferability due to the broad coverage of chemical space represented in their training sets, their size and computational cost limits their routine use. We address this gap by developing a lightweight general-purpose MLIAP for metal alloys spanning twelve elements (Al, Sc, Ti, V, Cu, Zn, Zr, Nb, Mo, Hf, Ta, W). Here, we utilize a step-wise active learning (AL) approach to develop a compact yet diverse dataset and model that generalizes well by interpolating from previously seen interactions. The resulting potentials transfer effectively to unseen, higher body-order alloys systems, requiring only unary and binary training data to capture ternary and quaternary systems. The model also transfers well to new elements via an alchemical extrapolation within the network, requiring only unary data for the new element and its alloys. These findings suggest that MLIAPs trained on targeted, diversity-driven datasets provide transferability at a relatively low computational cost, offering a practical path toward general-purpose, resource-efficient MLIAPs for materials modeling.
ER -