Metraipolis › archive › dc_08cc5367671b

SS-ESOAP: Self-Scaled Adaptive Preconditioning for Physics-Informed Learning — arXiv:2608.29448

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.29448
License (as archived)
CC0-1.0
Archived
2026-10-17T07:47:45.232Z
SHA-256 of the archived text
050f75c88887fcec155c4dd75c0b0d351b9375cd46e126ef9f97be4e7a444aa0
Stable URL of this page
https://metraipolis.com/documento/dc_08cc5367671b

How to cite

SS-ESOAP: Self-Scaled Adaptive Preconditioning for Physics-Informed Learning — arXiv:2608.29448 — arXiv (en), https://arxiv.org/abs/2608.29448 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_08cc5367671b (Metraipolis archive dc_08cc5367671b, sha256 050f75c88887fcec155c4dd75c0b0d351b9375cd46e126ef9f97be4e7a444aa0)
BibTeX
@misc{dc08cc5367671b,
  title = {SS-ESOAP: Self-Scaled Adaptive Preconditioning for Physics-Informed Learning — arXiv:2608.29448},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.29448},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_08cc5367671b; sha256 050f75c88887fcec155c4dd75c0b0d351b9375cd46e126ef9f97be4e7a444aa0},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.29448. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Physics-informed neural networks (PINNs) often face ill-conditioned objectives that limit high-accuracy training. Dense quasi-Newton methods improve local conditioning but require expensive optimizer state, while Kronecker-factored methods such as SOAP scale to larger networks but rely on periodic basis updates. We introduce \method, which augments SOAP-style preconditioning with a scalar secant-energy correction adapted to Kronecker geometry and an adaptive basis update followed by variance-state downscaling. We characterize the directional secant matching induced by the scalar correction and give a bound on variance-state mismatch across basis changes. Across eight PDE benchmarks, \method attains the lowest final residual on six, including Burgers and Boussinesq, while SOAP-family baselines perform better on Gray-Scott and Ginzburg-Landau. On Boussinesq, \method reaches a residual of $10^{-5}$ in 4.1 hours with 9.2 GB peak VRAM, while Adam does not reach this target within 14 hours. Three-seed $L^2$ and $H^1$ errors on four representative PDEs support the link between lower residuals and improved solution accuracy. These results position \method as a scalable option for stiff, high-accuracy physics-informed training, rather than a uniform replacement for existing optimizers.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-01 in cs.AI).
Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI); Optimization and Control (math.OC); Machine Learning (stat.ML).
Pagina del lavoro: https://arxiv.org/abs/2608.29448 — 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 2026-10-17T07:28:26.760Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).