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

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 Schrödinger 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.

References in corpus (6)

• Understanding molecular representations in machine learning: The role of uniqueness and target similarity
• $\Delta$-Machine Learning for Potential Energy Surfaces: A PIP approach to bring a DFT-based PES to CCSD(T) Level of Theory
• Multi-Determinant Wave-functions in Quantum Monte Carlo
• Informing Geometric Deep Learning with Electronic Interactions to Accelerate Quantum Chemistry
• Electronic Origin of the Volume Collapse in Cerium
• Bias Cancellation in One-Determinant Fixed-Node Diffusion Monte Carlo: Insights from Fermionic Occupation Numbers

Cited by in corpus (7)

• Towards self-driving laboratories: The central role of density functional theory in the AI age
• Accurate quantum Monte Carlo forces for machine-learned force fields: Ethanol as a benchmark
• Accurate and efficient machine learning interatomic potentials for finite temperature modeling of molecular crystals
• Basis set incompleteness errors in fixed-node diffusion Monte Carlo calculations on non-covalent interactions
• Reproducibility of fixed-node diffusion Monte Carlo across diverse community codes: The case of water-methane dimer
• Evolutionary Monte Carlo of QM properties in chemical space: Electrolyte design
• Gaussian Processes for Finite Size Extrapolation of Many-Body Simulations