Saved in:
Bibliographic Details
Main Authors: Alcalde, Albert, Widhalm, Markus, Yılmaz, Emre
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2602.08478
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912890451132416
author Alcalde, Albert
Widhalm, Markus
Yılmaz, Emre
author_facet Alcalde, Albert
Widhalm, Markus
Yılmaz, Emre
contents We propose the time-delayed transformer (TD-TF), a simplified transformer architecture for data-driven modeling of unsteady spatio-temporal dynamics. TD-TF bridges linear operator-based methods and deep sequence models by showing that a single-layer, single-head transformer can be interpreted as a nonlinear generalization of time-delayed dynamic mode decomposition (TD-DMD). The architecture is deliberately minimal, consisting of one self-attention layer with a single query per prediction and one feedforward layer, resulting in linear computational complexity in sequence length and a small parameter count. Numerical experiments demonstrate that TD-TF matches the performance of strong linear baselines on near-linear systems, while significantly outperforming them in nonlinear and chaotic regimes, where it accurately captures long-term dynamics. Validation studies on synthetic signals, unsteady aerodynamics, the Lorenz '63 system, and a reaction-diffusion model show that TD-TF preserves the interpretability and efficiency of linear models while providing substantially enhanced expressive power for complex dynamics.
format Preprint
id arxiv_https___arxiv_org_abs_2602_08478
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Time-Delayed Transformers for Data-Driven Modeling of Low-Dimensional Dynamics
Alcalde, Albert
Widhalm, Markus
Yılmaz, Emre
Machine Learning
Numerical Analysis
Dynamical Systems
37M10 (Primary) 68T07, 37N30 (Secondary)
We propose the time-delayed transformer (TD-TF), a simplified transformer architecture for data-driven modeling of unsteady spatio-temporal dynamics. TD-TF bridges linear operator-based methods and deep sequence models by showing that a single-layer, single-head transformer can be interpreted as a nonlinear generalization of time-delayed dynamic mode decomposition (TD-DMD). The architecture is deliberately minimal, consisting of one self-attention layer with a single query per prediction and one feedforward layer, resulting in linear computational complexity in sequence length and a small parameter count. Numerical experiments demonstrate that TD-TF matches the performance of strong linear baselines on near-linear systems, while significantly outperforming them in nonlinear and chaotic regimes, where it accurately captures long-term dynamics. Validation studies on synthetic signals, unsteady aerodynamics, the Lorenz '63 system, and a reaction-diffusion model show that TD-TF preserves the interpretability and efficiency of linear models while providing substantially enhanced expressive power for complex dynamics.
title Time-Delayed Transformers for Data-Driven Modeling of Low-Dimensional Dynamics
topic Machine Learning
Numerical Analysis
Dynamical Systems
37M10 (Primary) 68T07, 37N30 (Secondary)
url https://arxiv.org/abs/2602.08478