Metraipolis › archive › dc_d4478437a6cd

AccelNet: Exact backward-compatible acceleration of polynomial angular descriptors through Cartesian moment factorization — arXiv:2608.03280

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.03280
License (as archived)
CC0-1.0
Archived
2026-10-31T13:12:58.460Z
SHA-256 of the archived text
3d19918595da942898cce6d766f1cc6cee0055d5b6654c92788e325445a5cffd
Stable URL of this page
https://metraipolis.com/documento/dc_d4478437a6cd

How to cite

AccelNet: Exact backward-compatible acceleration of polynomial angular descriptors through Cartesian moment factorization — arXiv:2608.03280 — arXiv (en), https://arxiv.org/abs/2608.03280 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d4478437a6cd (Metraipolis archive dc_d4478437a6cd, sha256 3d19918595da942898cce6d766f1cc6cee0055d5b6654c92788e325445a5cffd)
BibTeX
@misc{dcd4478437a6cd,
  title = {AccelNet: Exact backward-compatible acceleration of polynomial angular descriptors through Cartesian moment factorization — arXiv:2608.03280},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.03280},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d4478437a6cd; sha256 3d19918595da942898cce6d766f1cc6cee0055d5b6654c92788e325445a5cffd},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.03280. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

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 $\cos \theta$, 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.

Tema degli esploratori: fisica e astronomia (retro-catalogo arXiv: annunci del 2026-08-05 in physics.comp-ph).
Categorie arXiv: Materials Science (cond-mat.mtrl-sci) ; Computational Physics (physics.comp-ph).
Pagina del lavoro: https://arxiv.org/abs/2608.03280 — 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 2026-10-31T13:01:51.230Z dalla pagina https://arxiv.org/catchup/physics.comp-ph/2026-08-05?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).