Metraipolis › archive › dc_2d80693423dd
Predicting Out-of-Distribution Generalization of Neural Operators via Observable Spectral Error Decomposition — arXiv:2609.23529
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.23529
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-09T07:07:14.642Z
- SHA-256 of the archived text
9f8695b2c268b9badbb2fe8a7c68fa55c92c2130425fbb10db1e345b28d47437- Stable URL of this page
https://metraipolis.com/documento/dc_2d80693423dd
How to cite
Predicting Out-of-Distribution Generalization of Neural Operators via Observable Spectral Error Decomposition — arXiv:2609.23529 — arXiv (en), https://arxiv.org/abs/2609.23529 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2d80693423dd (Metraipolis archive dc_2d80693423dd, sha256 9f8695b2c268b9badbb2fe8a7c68fa55c92c2130425fbb10db1e345b28d47437)
BibTeX
@misc{dc2d80693423dd,
title = {Predicting Out-of-Distribution Generalization of Neural Operators via Observable Spectral Error Decomposition — arXiv:2609.23529},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.23529},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2d80693423dd; sha256 9f8695b2c268b9badbb2fe8a7c68fa55c92c2130425fbb10db1e345b28d47437},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.23529. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Neural operators have emerged as powerful surrogates for solving partial differential equations (PDEs), yet their reliability under distribution shift remains a critical barrier to deployment. Existing approaches to out-of-distribution (OOD) generalization in operator learning are largely empirical and black-box: they report aggregate error metrics without explaining why errors arise or when they will grow. We propose a structure-preserving framework that makes OOD generalization predictable and auditable. Our key idea is to parameterize the learned solution operator as a spectral filter $h_\theta(\lambda)$ acting on the eigenvalues of the underlying elliptic operator, implemented via Chebyshev polynomial expansions and trained with a weak-form objective. This parameterization admits an exact decomposition of the energy-norm error into two observable components: a model-dependent spectral approximation term and a distribution-dependent spectral weighting term induced by the input. From this decomposition we derive three diagnostics: a conservative in-band supremum $\vareps_{\mathrm{sup}}$, a global RMS proxy $\vareps_{\mathrm{rms}}$, and a sample-dependent effective metric $\vareps_{\mathrm{eff}}(f)$. These diagnostics can be computed without access to ground-truth solutions. Through four controlled experiments, we show that $\vareps_{\mathrm{eff}}(f)\|f\|$ consistently predicts energy error under in-distribution, in-band spectral shift, out-of-band tail, and compound shifts, whereas global metrics can be systematically misleading. Our framework shifts OOD assessment of neural operators from black-box benchmarking to operator-structure diagnostics, providing a practical route to auditable scientific machine learning.
Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-22 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.23529 — 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 2026-10-09T06:58:09.307Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-22?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).