Amortized Generative Modeling of Invariant Measures across A Family of Kuramoto-Sivashinsky PDEs
Abstract
The long-run statistics of chaotic dissipative partial differential equations are described by invariant measures, and when the equation carries a parameter these measures form a family whose members can differ in kind, from steady states to sustained spatiotemporal chaos. We ask whether a single generative model can cover such a family. We train a single conditional diffusion model once, on states of the Kuramoto-Sivashinsky equation at 160 values of its hyperviscosity, and evaluate it against held-out data at interleaved parameter values not seen in training, using a classifier two-sample test supported by spectral, geometric and tail statistics. Over most of the family, agreement improves as the dynamics become more chaotic. We identify a structural cause: the weakly chaotic members have invariant measures concentrated on low-dimensional sets, which a model with full support output cannot represent, and the target approaches full dimensionality as chaos increases. Restricting training to the strongly chaotic members reduces the classifier's excess over chance on those members by roughly 40%, and this restricted conditional model matches or exceeds models trained at single parameter values. On those members, generated samples reproduce energy spectra to within a resampling floor, with a small systematic deficit at the largest scales, and match the energy balance of held-out data.
Related
- Whole-brain modeling of dynamic causal circuits in human cognition using amortized variational inference
- Stable invariant measures in linear dynamics
- Geometric structures, orbit closures and invariant measures in moduli spaces
- Absolute continuity of invariant measures for random diffeomorphisms
- Scale-invariant bounds on thin sets for measures satisfying a first-order PDE and the anisotropic Michael-Simon inequality
BibTeX
@misc{roy2026amortized,
title = {{Amortized Generative Modeling of Invariant Measures across A Family of Kuramoto-Sivashinsky PDEs}},
author = {Arnab Roy and Tobin A. Driscoll},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.32914},
eprint = {2609.32914},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.32914},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Amortized Generative Modeling of Invariant Measures across A Family of Kuramoto-Sivashinsky PDEs
AU - Arnab Roy
AU - Tobin A. Driscoll
PY - 2026
DA - 2026/09/26
PB - arXiv
DO - 10.48550/arxiv.2609.32914
UR - https://arxiv.org/abs/2609.32914
AB - The long-run statistics of chaotic dissipative partial differential equations are described by invariant measures, and when the equation carries a parameter these measures form a family whose members can differ in kind, from steady states to sustained spatiotemporal chaos. We ask whether a single generative model can cover such a family. We train a single conditional diffusion model once, on states of the Kuramoto-Sivashinsky equation at 160 values of its hyperviscosity, and evaluate it against held-out data at interleaved parameter values not seen in training, using a classifier two-sample test supported by spectral, geometric and tail statistics. Over most of the family, agreement improves as the dynamics become more chaotic. We identify a structural cause: the weakly chaotic members have invariant measures concentrated on low-dimensional sets, which a model with full support output cannot represent, and the target approaches full dimensionality as chaos increases. Restricting training to the strongly chaotic members reduces the classifier's excess over chance on those members by roughly 40%, and this restricted conditional model matches or exceeds models trained at single parameter values. On those members, generated samples reproduce energy spectra to within a resampling floor, with a small systematic deficit at the largest scales, and match the energy balance of held-out data.
ER -