LATTE: an atomic environment descriptor based on Cartesian tensor contractions

TL;DR

LATTE is a new atomic environment descriptor based on Cartesian tensor contractions that enables efficient, scalable, and expressive machine learning potentials with adjustable accuracy and computational cost.

Contribution

The paper introduces LATTE, a novel descriptor using tensor contractions that improves flexibility and efficiency in constructing interatomic potentials for machine learning models.

Findings

Competitive with fast potentials

Extensible to near state-of-the-art accuracy

Efficient for various system sizes

Abstract

We propose a new descriptor for local atomic environments, to be used in combination with machine learning models for the construction of interatomic potentials. The Local Atomic Tensors Trainable Expansion (LATTE) allows for the efficient construction of a variable number of many-body terms with learnable parameters, resulting in a descriptor that is efficient, expressive, and can be scaled to suit different accuracy and computational cost requirements. We compare this new descriptor to existing ones on several systems, showing it to be competitive with very fast potentials at one end of the spectrum, and extensible to an accuracy close to the state of the art.