{
  "abstract": "We present AccelNet, an exact, backward-compatible method for accelerating existing trained aenet and n2p2 neural-network potentials without retraining. For angular terms with separable one-neighbor weights and a finite polynomial dependence on $\\cos θ$, the method exploits their hidden finite-rank structure to replace explicit neighbor-pair loops by one-neighbor Cartesian moments. AccelNet reads models trained with either package and reproduces their descriptors, energies, and analytic forces to floating-point roundoff. We verified this equivalence for H$_2$O and TiO$_2$ models and tested the resulting potentials in LAMMPS molecular-dynamics simulations. The implementation, model-conversion tools, and LAMMPS interfaces are released as open-source software. AccelNet also enables GPU-accelerated molecular dynamics with existing aenet and n2p2 potentials. Moment evaluation achieves speedups of up to 10.9 on a CPU and 15.8 on a GPU relative to direct evaluation within AccelNet. The implementation, model-conversion tools, and LAMMPS interfaces are released as open-source software.",
  "abstract_shown": true,
  "affiliations": [],
  "article_type": null,
  "arxiv": "2608.03280",
  "author_total": 1,
  "authors": [
    {
      "name": "Yuki Nagai"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/4ezxzqh4eavgv.bib",
  "cited_by": 0,
  "doi": "10.48550/arxiv.2608.03280",
  "fulltext_url": null,
  "id": "4ezxzqh4eavgv",
  "issns": [],
  "landing_url": "https://arxiv.org/abs/2608.03280",
  "language": null,
  "licence": "LicenseRef-arXiv-nonexclusive-1.0",
  "licence_label": "arXiv licence",
  "licence_source": "article",
  "licence_url": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
  "metadata_licence": "CC0-1.0",
  "native_subjects": [
    [
      "arxiv",
      "cond-mat.mtrl-sci"
    ],
    [
      "arxiv",
      "physics.comp-ph"
    ]
  ],
  "openalex": null,
  "pdf_url": "https://arxiv.org/pdf/2608.03280",
  "pmcid": null,
  "pmid": null,
  "published": "2026-09-30",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/4ezxzqh4eavgv.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:25:48Z",
      "native_id": "2608.03280",
      "source": "arxiv",
      "url": "https://arxiv.org/abs/2608.03280"
    }
  ],
  "subjects": [
    {
      "field": "physics",
      "origin": "arxiv:physics.comp-ph",
      "subfield": "computational"
    },
    {
      "field": "physics",
      "origin": "arxiv:cond-mat.mtrl-sci",
      "subfield": "condensed-matter"
    }
  ],
  "title": "AccelNet: Exact backward-compatible acceleration of polynomial angular descriptors through Cartesian moment factorization",
  "title_text": "AccelNet: Exact backward-compatible acceleration of polynomial angular descriptors through Cartesian moment factorization",
  "type": "preprint",
  "updated_at": "2026-10-30T12:25:48Z",
  "url": "https://research.draconitic.com/paper/4ezxzqh4eavgv",
  "venue": "arXiv",
  "versions": []
}