Metraipolis › archive › dc_1d7616ef3483

When Known Physics Helps Neural PDE Models: Residual Constraints Out-Regularize Generic Priors for Nonlinear Dynamics — arXiv:2609.34012

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.34012
License (as archived)
CC0-1.0
Archived
2026-10-07T21:02:00.017Z
SHA-256 of the archived text
d93df596d74a9f78b89d4eee9747034a90cab7b901bb7325086ae4a76dab6172
Stable URL of this page
https://metraipolis.com/documento/dc_1d7616ef3483

How to cite

When Known Physics Helps Neural PDE Models: Residual Constraints Out-Regularize Generic Priors for Nonlinear Dynamics — arXiv:2609.34012 — arXiv (en), https://arxiv.org/abs/2609.34012 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_1d7616ef3483 (Metraipolis archive dc_1d7616ef3483, sha256 d93df596d74a9f78b89d4eee9747034a90cab7b901bb7325086ae4a76dab6172)
BibTeX
@misc{dc1d7616ef3483,
  title = {When Known Physics Helps Neural PDE Models: Residual Constraints Out-Regularize Generic Priors for Nonlinear Dynamics — arXiv:2609.34012},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.34012},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_1d7616ef3483; sha256 d93df596d74a9f78b89d4eee9747034a90cab7b901bb7325086ae4a76dab6172},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.34012. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Neural PDE surrogates increasingly incorporate structural priors, yet it is often unclear whether their gains arise from physics-specific information or simply from regularization and training choices. We evaluate several such priors under a common protocol against a matched from-scratch neural operator baseline. Our central result is that a known-equation residual consistently outperforms the best generic regularizer at equal tuning budget. At fixed capacity this benefit appears across linear and nonlinear PDEs, but a capacity sweep reveals a sharp distinction: the advantage persists and grows for Burgers, KdV, and Allen-Cahn, while collapsing toward or below parity for linear heat and advection-diffusion. Thus, the durable value of the residual is specific to nonlinear operators. We further falsify a pre-registered hypothesis that the benefit is activated only by data sparsity: the residual remains advantageous even under full supervision. Its usefulness does, however, have a clear boundary. Under grid under-resolution, nonlinear coarse fields no longer satisfy the naive governing-equation residual, and enforcing it becomes actively harmful. In contrast, cross-family pretraining and in-context conditioning fail to outperform the strong from-scratch baseline in the regime studied. Together, these results identify when known physics provides non-redundant information to neural PDE models, when it does not, and when enforcing it introduces bias.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.AI).
Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.34012 — 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 2026-10-07T20:29:02.814Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).