Dual Benchmarking of Universal Machine Learning Interatomic Potentials Against DFT and Experimental EXAFS Data

Pjotrs Žguns, Inga Pudza, Alexei Kuzmin

Keywords: universal machine learning interatomic potentials, uMLIPs, validation, benchmarking, experimental verification, EXAFS, structure

TL;DR: Benchmarking uMLIPs against both ab initio and experimental data enhances confidence in their reliability and confirms their experimental-level accuracy for materials modeling.

Confirmation Of Submission Requirements: I submit an abstract. It uses the template provided on the submission page and is no longer than 2 pages.

Submission Number: 234