Building useful MLIPs
Gus Hart, Brigham Young University, USA
Machine-learned interatomic potentials are far more expressive than traditional physically motivated interatomic potentials like Lennard-Jones, Stillinger-Weber, Embedded Atom Potentials, etc. While they can be far more accurate, they are also more likely to be completely wrong outside of the training domain, are more difficult to train reliably, and are computationally expensive. We have developed MLIPs for the Hf-Ni-Ti shape memory alloy. We share cautionary tales, best practices for generating training sets, and demonstrate how community tools make for “easy entry” to realistic thermodynamic modeling with these potentials.