Gaussian Plane-Wave Neural Operator for Electron Density Estimation

TL;DR

This paper introduces the Gaussian plane-wave neural operator (GPWNO), a novel machine learning model that effectively predicts electron densities by operating in functional space with plane-wave and Gaussian bases, outperforming existing methods.

Contribution

The paper presents GPWNO, a new neural operator that combines plane-wave and Gaussian bases for improved electron density estimation in DFT-related tasks.

Findings

GPWNO outperforms ten baseline models on multiple datasets.

Effective representation of both high- and low-frequency density components.

Demonstrates superior accuracy in electron density prediction.

Abstract

This work studies machine learning for electron density prediction, which is fundamental for understanding chemical systems and density functional theory (DFT) simulations. To this end, we introduce the Gaussian plane-wave neural operator (GPWNO), which operates in the infinite-dimensional functional space using the plane-wave and Gaussian-type orbital bases, widely recognized in the context of DFT. In particular, both high- and low-frequency components of the density can be effectively represented due to the complementary nature of the two bases. Extensive experiments on QM9, MD, and material project datasets demonstrate GPWNO's superior performance over ten baselines.