metadata
license: apache-2.0
tags:
- turbulence
- Chaos
- Vorticity
pretty_name: tcf_full
We present the dataset accompanying the ICML 2025 paper “Chaos Meets Attention: Transformers for Large-Scale Dynamical Prediction.” In this work, we proposed a novel architecture of attention for spatial correlation in modelling high-dimensional chaotic systems, with a focus on achieving faithful long-time statistical behaviour.
This dataset addresses a benchmark on the Turbulent Channel Flow.