Normalizing flows for atomic solids
This work addresses the challenge of efficiently sampling and estimating free energies for atomic solids in computational physics, offering a method that eliminates the need for multi-staging, though it is incremental as it applies an existing technique to a specific domain.
The researchers tackled the problem of modeling atomic solids by developing a machine-learning approach using normalizing flows, which transforms a base distribution into target solids without ground-truth training samples, resulting in Helmholtz free energy estimates in excellent agreement with literature values and baseline methods, and samples nearly indistinguishable from molecular dynamics.
We present a machine-learning approach, based on normalizing flows, for modelling atomic solids. Our model transforms an analytically tractable base distribution into the target solid without requiring ground-truth samples for training. We report Helmholtz free energy estimates for cubic and hexagonal ice modelled as monatomic water as well as for a truncated and shifted Lennard-Jones system, and find them to be in excellent agreement with literature values and with estimates from established baseline methods. We further investigate structural properties and show that the model samples are nearly indistinguishable from the ones obtained with molecular dynamics. Our results thus demonstrate that normalizing flows can provide high-quality samples and free energy estimates without the need for multi-staging.