FrOGS: Discrete Neural Sampler for Independent Alloy Configurations Across Chemical Conditions
Abstract
Predicting the thermodynamic properties of an alloy requires sampling its configurations across many chemical conditions and recovering free energies on a common absolute scale. Markov chain Monte Carlo (MCMC) is the standard tool, but it requires separate simulations at different conditions, and auxiliary free-energy methods such as thermodynamic integration are used to place results on a common absolute scale. Modern discrete neural samplers typically use reverse KL divergence as the objective and can be mode-seeking or biased. We present Free energy Offering Generative Sampler (FrOGS), a hybrid discrete neural sampler that couples an autoregressive model to a continuous-time Markov chain (CTMC) to be trained jointly under a single shared loss. FrOGS draws i.i.d. configurations, returns an unbiased estimate of the partition function, and gives consistent estimates of thermodynamic observables. We train a single model across a wide range of chemical conditions to produce estimates on a common absolute free-energy scale. FrOGS matches exact finite-size results on the 2D Ising model and reference phase diagrams for AgPd and CuAu, without mode collapse. We additionally compare to SEGAL, a published autoregressive baseline, and find that only FrOGS recovers the stability range of the CuAu3 phase.
Related
- Existence and Uniqueness of Nearest Stealthy Hyperuniform Configurations from Random Initial Conditions
- An Affine Semigroup from Orbifold Boundary Conditions: cut, phylogenetic and hierarchical models in the unit-weight sector, and weighted configurations beyond them
- SU(9) grand unified model with rank-reducing discrete boundary conditions on T2/ℤ4
- Discrete Laplacian structure and kernel reduction of the gap equation in odd-dimensional large-N vector models at imaginary chemical potential
- Influence of chemical vapor deposition conditions on N incorporation ratio on vicinal 4H-SiC(000-1) surface: Ab Initio-based approach
BibTeX
@misc{min2026frogs,
title = {{FrOGS: Discrete Neural Sampler for Independent Alloy Configurations Across Chemical Conditions}},
author = {Kyucheol Min and Elyssa Hofgard and Tess Smidt},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.02948},
eprint = {2609.02948},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.02948},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - FrOGS: Discrete Neural Sampler for Independent Alloy Configurations Across Chemical Conditions
AU - Kyucheol Min
AU - Elyssa Hofgard
AU - Tess Smidt
PY - 2026
DA - 2026/09/21
PB - arXiv
DO - 10.48550/arxiv.2609.02948
UR - https://arxiv.org/abs/2609.02948
AB - Predicting the thermodynamic properties of an alloy requires sampling its configurations across many chemical conditions and recovering free energies on a common absolute scale. Markov chain Monte Carlo (MCMC) is the standard tool, but it requires separate simulations at different conditions, and auxiliary free-energy methods such as thermodynamic integration are used to place results on a common absolute scale. Modern discrete neural samplers typically use reverse KL divergence as the objective and can be mode-seeking or biased. We present Free energy Offering Generative Sampler (FrOGS), a hybrid discrete neural sampler that couples an autoregressive model to a continuous-time Markov chain (CTMC) to be trained jointly under a single shared loss. FrOGS draws i.i.d. configurations, returns an unbiased estimate of the partition function, and gives consistent estimates of thermodynamic observables. We train a single model across a wide range of chemical conditions to produce estimates on a common absolute free-energy scale. FrOGS matches exact finite-size results on the 2D Ising model and reference phase diagrams for AgPd and CuAu, without mode collapse. We additionally compare to SEGAL, a published autoregressive baseline, and find that only FrOGS recovers the stability range of the CuAu_3 phase.
ER -