Latent Dynamics · 2026-11-10

TORNADO: Adaptive Latent Stochastic Transport for Calibrated Probabilistic PDE Forecasting

Jorge Mifsut-Benet, Armand Kassa\"i Koupa\"i, Ramon Daniel Regueiro-Espino, Nicolas Baskiotis, Patrick Gallinari

arXiv:2611.10251PDF

AI-assisted summary based on the paper. Consult the linked source for methods and limitations. Benchmark numbers appear after editor verification.

TL;DR

TORNADO is a latent-space probabilistic PDE forecasting framework that models the next-step as stochastic transport between VAE encoder posteriors using a learned Itô SDE. Its key idea is state-dependent diffusion (learned drift and diffusion) to adapt predictive uncertainty to local latent dynamics. Training uses simulation-free analytic conditional endpoint targets from a Gaussian interpolant plus an energy-distance regularizer.

Why it matters

The authors target miscalibration of generative probabilistic surrogates in compute-constrained, few-step inference regimes, where fixed (state-agnostic) stochasticity can lead to overconfident forecasts. By learning an SDE with state-dependent diffusion in latent space and supervising it with analytic endpoint targets, TORNADO aims to improve uncertainty calibration while keeping predictive accuracy competitive.

Method

  • Model next-step latent forecasting as stochastic transport between consecutive VAE encoder posteriors using a learned latent Itô SDE.
  • Learn state-dependent diffusion (not a global schedule) so predictive uncertainty adapts to local latent dynamical regimes.
  • Train drift and diffusion with simulation-free analytic conditional endpoint targets from a Gaussian interpolant; add energy-distance regularization for distributional alignment.
Abstract (from arXiv)

Probabilistic neural surrogates for PDEs enable ensemble forecasting, but often exhibit miscalibration when used for uncertainty quantification, especially in compute-constrained, few-step inference regimes. Existing generative approaches typically rely on fixed global stochasticity, which cannot adapt predictive uncertainty to the local dynamical state. We propose TORNADO, a latent probabilistic forecasting framework that formulates next-step prediction as stochastic transport between VAE encoder posteriors. Unlike existing approaches with globally prescribed stochasticity, TORNADO learns a SDE with state-dependent diffusion, enabling the predictive uncertainty to adapt to the local latent dynamics. The model is trained using analytic conditional endpoint targets derived from a Gaussian interpolant, enabling simulation-free learning. An additional energy distance regularization term improves distributional alignment. Empirically, TORNADO improves uncertainty calibration over representative generative baselines, including stochastic interpolants, diffusion models, and flow matching, while maintaining comparable predictive accuracy. Our results show that adapting stochasticity to the latent state, rather than prescribing it globally, is an effective mechanism for improving calibration in practical neural PDE forecasting settings. Project page: https://tornado-sde.github.io.

Related papers

Publication record

Bibliographic evidence describes where and when a record appeared. It does not validate scientific claims.

First announced by arXiv: 2026-11-10 · Indexed by Pub-AI: 2026-11-10 UTC

preprint

TORNADO: Adaptive Latent Stochastic Transport for Calibrated Probabilistic PDE Forecasting

Jorge Mifsut Benet, Armand Kassa\"i Koupa\"i, Ramon Daniel Regueiro-Espino, Nicolas Baskiotis, Patrick Gallinari

First announced by arXiv: 2026-11-10 · Original record ↗

arxiv
2611.10251
Explicit source versions (1)
  • v1 · Date not supplied
  1. Preprint available

    Source date: 2026-11-10 · First detected: 2026-11-10 UTC · Historical baseline

    Source evidence (1)
    • Official evidence ↗ · RSS guid=oai:arXiv.org:2611.10251v1; arxiv:announce_type=new; item pubDate (official metadata)
Abstract · arxiv

Probabilistic neural surrogates for PDEs enable ensemble forecasting, but often exhibit miscalibration when used for uncertainty quantification, especially in compute-constrained, few-step inference regimes. Existing generative approaches typically rely on fixed global stochasticity, which cannot adapt predictive uncertainty to the local dynamical state. We propose TORNADO, a latent probabilistic forecasting framework that formulates next-step prediction as stochastic transport between VAE encoder posteriors. Unlike existing approaches with globally prescribed stochasticity, TORNADO learns a SDE with state-dependent diffusion, enabling the predictive uncertainty to adapt to the local latent dynamics. The model is trained using analytic conditional endpoint targets derived from a Gaussian interpolant, enabling simulation-free learning. An additional energy distance regularization term improves distributional alignment. Empirically, TORNADO improves uncertainty calibration over representative generative baselines, including stochastic interpolants, diffusion models, and flow matching, while maintaining comparable predictive accuracy. Our results show that adapting stochasticity to the latent state, rather than prescribing it globally, is an effective mechanism for improving calibration in practical neural PDE forecasting settings. Project page: https://tornado-sde.github.io.