{
  "annotation_id": "f171fd9b-576a-4bb1-b1f8-61dab4d4ba58",
  "date_created": "2026-02-17T05:53:15.893000Z",
  "date_modified": "2026-02-17T05:53:15.893000Z",
  "file_hash": "3be06ecb790ca63591d90a1cd41adf6849f92aef8671683d707b0b3a31fa28fc",
  "private": false,
  "record": {
    "abstract": "A multi-preconditioned LBFGS (MP-LBFGS) algorithm is introduced for training finite-basis physics-informed neural networks (FBPINNs). The algorithm is motivated by the nonlinear additive Schwarz method and exploits the domain-decomposition-inspired additive architecture of FBPINNs, in which local neural networks are defined on subdomains, thereby localizing the network representation. Parallel, subdomain-local quasi-Newton corrections are then constructed on the corresponding local parts of the architecture. A key feature is a novel nonlinear multi-preconditioning mechanism, in which subdomain corrections are optimally combined through the solution of a low-dimensional subspace minimization problem. Numerical experiments indicate that MP-LBFGS can improve convergence speed, as well as model accuracy over standard LBFGS while incurring lower communication overhead.",
    "arxiv_id": "2601.08709",
    "authors": [
      "Marc Salvad\u00f3-Benasco",
      "Aymane Kssim",
      "Alexander Heinlein",
      "Rolf Krause",
      "Serge Gratton",
      "Alena Kopani\u010d\u00e1kov\u00e1"
    ],
    "categories": [
      "math.NA",
      "cs.LG",
      "cs.NA"
    ],
    "license": "http://creativecommons.org/licenses/by-nc-nd/4.0/",
    "title": "Multi-Preconditioned LBFGS for Training Finite-Basis PINNs",
    "url": "https://arxiv.org/abs/2601.08709",
    "version": "v1"
  },
  "schema_id": "dorsal/arxiv",
  "source": {
    "execution_id": "bf9e0234-d905-4055-9acb-60872acbf5ad",
    "id": "arXiv Dataset",
    "type": "Model",
    "variant": "snapshot-2026-01-17",
    "version": "0.1.0"
  },
  "user_id": 1000002
}