Metraipolis › archive › dc_76ecb134fb7e
Learnable composition for neural operators — arXiv:2609.03069
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.03069
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T02:07:44.116Z
- SHA-256 of the archived text
e0f99d4748d28ec9a0a08658269e2f5f741d45922f6f7712312670d905b47211- Stable URL of this page
https://metraipolis.com/documento/dc_76ecb134fb7e
How to cite
Learnable composition for neural operators — arXiv:2609.03069 — arXiv (en), https://arxiv.org/abs/2609.03069 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_76ecb134fb7e (Metraipolis archive dc_76ecb134fb7e, sha256 e0f99d4748d28ec9a0a08658269e2f5f741d45922f6f7712312670d905b47211)
BibTeX
@misc{dc76ecb134fb7e,
title = {Learnable composition for neural operators — arXiv:2609.03069},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.03069},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_76ecb134fb7e; sha256 e0f99d4748d28ec9a0a08658269e2f5f741d45922f6f7712312670d905b47211},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.03069. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Neural operators are fast, differentiable surrogates for physical simulation, but their accuracy often degrades when domain geometry, size, or operating conditions differ from training. Supervised adaptation can recover accuracy, but even a small target set requires costly high-fidelity simulations. We therefore ask how pretraining and transfer can be designed together to reduce this deployment cost. LatentDDM first pretrains a neural operator to predict fields on small subdomains. For a new setting, it freezes this operator and trains only a lightweight module that composes the local predictions. We evaluate our method on two complementary problems: steady Darcy flow, where long-range pressure coupling must extend across increasingly large porous domains, and unsteady incompressible flow around a pitching airfoil, where rollout errors compound as target pitching frequencies exceed the training range. Compared with the capacity-matched models that process the full domain at once, LatentDDM's error is 36-56% lower on larger Darcy domains after adaptation with 16 target simulations. It also improves 20-step field rollouts in fast-pitching airfoil flow, both zero-shot and after few-shot calibration. These results identify the co-designed local pretraining and composition-level transfer as a promising design principle for physical foundation models. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-04 in cs.LG). Categorie arXiv: Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2609.03069 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-17T01:47:25.518Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-04?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).