English

Training-Free Bayesian Filtering with Generative Emulators

Machine Learning 2026-05-20 v1 Atmospheric and Oceanic Physics

Abstract

Bayesian filtering is a well-known problem that aims to estimate plausible states of a dynamical system from observations. Among existing approaches to solve this problem, particle filters are theoretically exact for non-linear dynamics and observations, but suffer from poor scalability in high dimensions. In this work, we show that diffusion-based emulators of dynamical systems can be used to implement, without additional training, an optimal variant of particle filters that has remained largely unexplored due to implementation challenges with classical numerical solvers. Experiments on nonlinear chaotic systems, including atmospheric dynamics, demonstrate that the proposed approach successfully scales particle filtering to high-dimensional settings.

Keywords

Cite

@article{arxiv.2605.20028,
  title  = {Training-Free Bayesian Filtering with Generative Emulators},
  author = {Thomas Savary and François Rozet and Gilles Louppe},
  journal= {arXiv preprint arXiv:2605.20028},
  year   = {2026}
}

Comments

Accepted as a spotlight paper at the International Conference on Machine Learning 2026