Bounding free energy difference with flow matching
arXiv:2311.07963 · doi:10.1088/0256-307X/40/12/120201
Abstract
This paper introduces a method for computing the Helmholtz free energy using the flow matching technique. Unlike previous work that utilized flow-based models for variational free energy calculations, this method provides bounds for free energy estimation based on targeted free energy perturbation, by performing calculations on samples from both ends of the mapping. We demonstrate applications of the present method by estimating the free energy of the classical Coulomb gas in a harmonic trap.
References in corpus (12)
- A nonequilibrium equality for free energy differences
- Machine learning and the physical sciences
- Path ensembles averages in systems driven far-from-equilibrium
- Neural Network Renormalization Group
- Asymptotically unbiased estimation of physical observables with neural samplers
- Targeted free energy perturbation
- Targeted free energy estimation via learned mappings
- Using bijective maps to improve free energy estimates
- Neural Canonical Transformation with Symplectic Flows
- Deep Variational Free Energy Approach to Dense Hydrogen
- Work Relations Connecting Nonequilibrium Steady States Without Detailed Balance
- of two-dimensional electron gas: a neural canonical transformation study