Hierarchical Implicit Neural Emulators
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866918046299324416 |
|---|---|
| author | Jiang, Ruoxi Zhang, Xiao Jakhar, Karan Lu, Peter Y. Hassanzadeh, Pedram Maire, Michael Willett, Rebecca |
| author_facet | Jiang, Ruoxi Zhang, Xiao Jakhar, Karan Lu, Peter Y. Hassanzadeh, Pedram Maire, Michael Willett, Rebecca |
| contents | Neural PDE solvers offer a powerful tool for modeling complex dynamical systems, but often struggle with error accumulation over long time horizons and maintaining stability and physical consistency. We introduce a multiscale implicit neural emulator that enhances long-term prediction accuracy by conditioning on a hierarchy of lower-dimensional future state representations. Drawing inspiration from the stability properties of numerical implicit time-stepping methods, our approach leverages predictions several steps ahead in time at increasing compression rates for next-timestep refinements. By actively adjusting the temporal downsampling ratios, our design enables the model to capture dynamics across multiple granularities and enforce long-range temporal coherence. Experiments on turbulent fluid dynamics show that our method achieves high short-term accuracy and produces long-term stable forecasts, significantly outperforming autoregressive baselines while adding minimal computational overhead. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_04528 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Hierarchical Implicit Neural Emulators Jiang, Ruoxi Zhang, Xiao Jakhar, Karan Lu, Peter Y. Hassanzadeh, Pedram Maire, Michael Willett, Rebecca Machine Learning Neural PDE solvers offer a powerful tool for modeling complex dynamical systems, but often struggle with error accumulation over long time horizons and maintaining stability and physical consistency. We introduce a multiscale implicit neural emulator that enhances long-term prediction accuracy by conditioning on a hierarchy of lower-dimensional future state representations. Drawing inspiration from the stability properties of numerical implicit time-stepping methods, our approach leverages predictions several steps ahead in time at increasing compression rates for next-timestep refinements. By actively adjusting the temporal downsampling ratios, our design enables the model to capture dynamics across multiple granularities and enforce long-range temporal coherence. Experiments on turbulent fluid dynamics show that our method achieves high short-term accuracy and produces long-term stable forecasts, significantly outperforming autoregressive baselines while adding minimal computational overhead. |
| title | Hierarchical Implicit Neural Emulators |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2506.04528 |