Towards DMC accuracy across chemical space with scalable $\Delta$-QML

TL;DR

This paper demonstrates that combining quantum diffusion Monte Carlo with scalable quantum machine learning models enables accurate predictions of molecular energetics across chemical space, significantly reducing computational costs.

Contribution

It introduces scalable $ riangle$-QML models that leverage minimal amons sets to efficiently predict energies with near chemical accuracy across diverse molecules.

Findings

DMC results for over 1,000 small organic molecules validate the approach.

Modest QMC training datasets suffice for accurate energy predictions.

$ riangle$-AQML models achieve near chemical accuracy across chemical space.

Abstract

In the past decade, quantum diffusion Monte Carlo (DMC) has been demonstrated to successfully predict the energetics and properties of a wide range of molecules and solids by numerically solving the electronic many-body Schr\"odinger equation. We show that when coupled with quantum machine learning (QML) based surrogate methods the computational burden can be alleviated such that QMC shows clear potential to undergird the formation of high quality descriptions across chemical space. We discuss three crucial approximations necessary to accomplish this: The fixed node approximation, universal and accurate references for chemical bond dissociation energies, and scalable minimal amons set based QML (AQML) models. Numerical evidence presented includes converged DMC results for over one thousand small organic molecules with up to 5 heavy atoms used as amons, and 50 medium sized organic molecules with 9 heavy atoms to validate the AQML predictions. Numerical evidence collected for $\Delta$-AQML models suggests that already modestly sized QMC training data sets of amons suffice to predict total energies with near chemical accuracy throughout chemical space.