Highly Accurate Real-space Electron Densities with Neural Networks
This work addresses a technical bottleneck in quantum chemistry for researchers needing accurate density-based observables, representing an incremental improvement with a novel hybrid approach.
The paper tackles the difficulty of extracting accurate electron densities from quantum chemical wave functions by introducing a neural network method that captures asymptotic properties and is trained via score matching and noise-contrastive estimation. It demonstrates highly accurate densities from deep QMC wave functions, enabling precise calculations of dipole moments, nuclear forces, and other properties.
Variational ab-initio methods in quantum chemistry stand out among other methods in providing direct access to the wave function. This allows in principle straightforward extraction of any other observable of interest, besides the energy, but in practice this extraction is often technically difficult and computationally impractical. Here, we consider the electron density as a central observable in quantum chemistry and introduce a novel method to obtain accurate densities from real-space many-electron wave functions by representing the density with a neural network that captures known asymptotic properties and is trained from the wave function by score matching and noise-contrastive estimation. We use variational quantum Monte Carlo with deep-learning ansätze (deep QMC) to obtain highly accurate wave functions free of basis set errors, and from them, using our novel method, correspondingly accurate electron densities, which we demonstrate by calculating dipole moments, nuclear forces, contact densities, and other density-based properties.