Towards Universal Neural Operators through Multiphysics Pretraining
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866909901060571136 |
|---|---|
| author | Masliaev, Mikhail Gusarov, Dmitry Markov, Ilya Hvatov, Alexander |
| author_facet | Masliaev, Mikhail Gusarov, Dmitry Markov, Ilya Hvatov, Alexander |
| contents | Although neural operators are widely used in data-driven physical simulations, their training remains computationally expensive. Recent advances address this issue via downstream learning, where a model pretrained on simpler problems is fine-tuned on more complex ones. In this research, we investigate transformer-based neural operators, which have previously been applied only to specific problems, in a more general transfer learning setting. We evaluate their performance across diverse PDE problems, including extrapolation to unseen parameters, incorporation of new variables, and transfer from multi-equation datasets. Our results demonstrate that advanced neural operator architectures can effectively transfer knowledge across PDE problems. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_10829 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Towards Universal Neural Operators through Multiphysics Pretraining Masliaev, Mikhail Gusarov, Dmitry Markov, Ilya Hvatov, Alexander Machine Learning Although neural operators are widely used in data-driven physical simulations, their training remains computationally expensive. Recent advances address this issue via downstream learning, where a model pretrained on simpler problems is fine-tuned on more complex ones. In this research, we investigate transformer-based neural operators, which have previously been applied only to specific problems, in a more general transfer learning setting. We evaluate their performance across diverse PDE problems, including extrapolation to unseen parameters, incorporation of new variables, and transfer from multi-equation datasets. Our results demonstrate that advanced neural operator architectures can effectively transfer knowledge across PDE problems. |
| title | Towards Universal Neural Operators through Multiphysics Pretraining |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2511.10829 |