#machine learning Preprint

Particle GFlowNets: Rethinking Generative Marginalization Models

Tiago da Silva Diego Mesquita S. Lahlou

Sep 2026 · 0 citations · 57 references

Computer Science

TL;DR

This work describes an automatic criterion for full-state rejuvenation of the Gibbs sampler, derived from the Gelman-Rubin statistic, which plays a key role in speeding up learning convergence.

Abstract

Generative Marginalization Models (MaMs) have been recently introduced as efficient neural sampling models for any-order autoregressive modelling of discrete distributions. By learning both the marginal and conditional probabilities of a persistent-block Gibbs sampler, MaMs enable fast posterior evaluation with a single neural network forward pass. While prior work has considered MaMs to be distinct from Generative Flow Networks (GFlowNets), a well-established paradigm for inference in discrete stochastic models, we show that they are equivalent. Then, we also extend MaMs'sampling strategy to non-autoregressive generative processes. In particular, we describe an automatic criterion for full-state rejuvenation of the Gibbs sampler, derived from the Gelman-Rubin statistic, which plays a key role in speeding up learning convergence. Our experiments show that our method, called Particle GFlowNets, markedly accelerates training in large combinatorial spaces.

DOI arXiv Semantic Scholar

View source

Similar papers

Preprint Aug 2026

Path-dependent Discrete Amortized Inference

It is demonstrated that the Markovian assumption can both hamper signal propagation during training and catastrophically reduce the learned sampler's expressivity due to state aliasing.

Tiago da Silva, Esmeralda S. Whitammer, S. Lahlou · 0 citations

Preprint Aug 2026

Parameter estimation in Conditional Sequential Monte Carlo algorithms through Particle Learning

The p(parameter)-CSMC algorithm is proposed, which incorporates both parameter learning and ancestor sampling, leading to much better mixing properties compared to (particle) Gibbs sampling in settings where strong internal correlations may challenge effective exploration.

A. D. Palomares, G. Storvik · 0 citations

#machine learning Preprint Sep 2026

Livin'on a Prior: Likelihood Score Approximation for Inverse Problems

Likelihood Score Approximation (LSA), a generative framework that keeps a pretrained unconditional model fixed and learns an observation-conditioned model that approximates the likelihood score from paired samples, is introduced.

Rostislav Makarov, Tal Peer, Danilo de Oliveira et al. · 0 citations

Preprint Aug 2026

Stochastic Autoregressive Learning

It is shown that stochastic autoregressive learning fundamentally differs from the deterministic theory, and that CoT learning at scale $\varepsilon$ is upper-bounded by base learning at scale $\varepsilon/M^2$, whereas e2e learning at scale $\varepsilon$ is upper-bounded, up to logarithmic factors, by $(M/\varepsilon)...

Ilan Doron-Arad, Idan Mehalel, Elchanan Mossel · 1 citation

#machine learning Preprint Sep 2026

Learning to Bias: Machine Learning-Enhanced Particle Filters

Neural Optimal Particle Filters are introduced, which learn an amortized approximation to the optimal proposal from offline simulated one-step conditioning tuples, and improve sample efficiency and distributional accuracy over standard PF baselines with modest computational overhead.

Apoorv Srivastava, Eric F. Darve · 0 citations

Review Aug 2026

Leveraging generative models to assist Monte Carlo sampling

A new paradigm that has recently emerged at the interface of machine learning and computational statistical physics: the use of generative models as tools for sampling through flexible probabilistic models that can assist the sampling of distributions known only up to a normalization constant is explored.

M. Gabrié · 0 citations