Grand Canonical Generators
cs.LG, stat.ML
Submitted: 2026-09-30
Updated: 2026-10-06
Comments: SimBioChem NeurIPS 206
License: http://creativecommons.org/licenses/by/4.0/
The gist: We introduce Grand Canonical Generators (GCG), a generative framework that extends Boltzmann generators to the grand canonical ensemble.
Terminology
Abstract
We introduce Grand Canonical Generators (GCG), a generative framework that extends Boltzmann generators to the grand canonical ensemble. We present two designs. The first conditions a variable-size generative model on the chemical potential, sampling particle number and configuration jointly. The second factorizes the grand canonical distribution into a particle-number distribution and the corresponding canonical Boltzmann density. This factorized formulation can use any existing Boltzmann generator for the canonical component, encodes the known linear chemical-potential dependence analytically, and yields a tractable likelihood that supports self-normalized importance sampling (SNIS). Empirically, GCG accurately reproduces grand canonical observables on a Lennard--Jones fluid and methane adsorption in a zeolite, demonstrating generalization across chemical potentials and correction via SNIS and grand canonical Monte Carlo.
Sources
- Shoot from the HIP: Hessian Interatomic Potentials without derivatives
- M\=oLe-{\Lambda}: Learning the Coupled-Cluster Response State for Energies, Gradients, and Properties
- Scalable Autoregressive 3D Molecule Generation
- Scaling Autoregressive Models for Lattice Thermodynamics
- Derivative Informed Learning of Exchange-Correlation Functionals
- Generative Molecular Morphing for Flexible-Size Design via Unbalanced Optimal Transport
- Boltzmann generators for amorphous particle systems
- Boltzmann Generators for Condensed Matter via Riemannian Flow Matching
- Grand canonical generative diffusion model for crystalline phases and grain boundaries
- Few-Step Boltzmann Generators via Scalable Likelihood Flow Maps
- Set Flow: A Permutation Invariant Normalizing Flow
- Autoregressive Boltzmann Generators
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks