Polynomial machine learning potential for Be-Ge developed by Seko (2022) v000

Read full text Summarise with VIRAT
Authors: Atsuto Seko
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Polynomial machine learning potential (MLP) for Be-Ge system. This potential is gtinv-269 potential in Be-Ge-2022-07-21 taken from Polynomial MLP Repository. The RMS errors for energy and forces are 2.754 meV/atom and 0.0601 eV/angstrom, respectively. The estimated computational cost required for a single core calculation is 1.46 ms/atom/step.

Fields of Study & Subjects

Machine Learning in Materials Science Advanced Chemical Physics Studies Thermodynamic and Structural Properties of Metals and Alloys

Publication details

Journal / source
Open Knowledgebase of Interatomic Models (OpenKIM)
Publication date
1 January 2026
Language
EN