Metraipolis › archive › dc_7bbba9e7bc3f

Truncated automatic sparse differentiation for machine learning interatomic potentials — arXiv:2609.20510

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.20510
License (as archived)
CC0-1.0
Archived
2026-10-10T04:09:26.661Z
SHA-256 of the archived text
a692db4cf4db2b2bf58ecacf3a27f844d5aa5f9231bd5c298fe463673f30438e
Stable URL of this page
https://metraipolis.com/documento/dc_7bbba9e7bc3f

How to cite

Truncated automatic sparse differentiation for machine learning interatomic potentials — arXiv:2609.20510 — arXiv (en), https://arxiv.org/abs/2609.20510 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7bbba9e7bc3f (Metraipolis archive dc_7bbba9e7bc3f, sha256 a692db4cf4db2b2bf58ecacf3a27f844d5aa5f9231bd5c298fe463673f30438e)
BibTeX
@misc{dc7bbba9e7bc3f,
  title = {Truncated automatic sparse differentiation for machine learning interatomic potentials — arXiv:2609.20510},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.20510},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7bbba9e7bc3f; sha256 a692db4cf4db2b2bf58ecacf3a27f844d5aa5f9231bd5c298fe463673f30438e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.20510. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Machine learning interatomic potentials (MLIPs) learn the mapping from atomic positions to potential energy. The forces, the negative gradient of this energy, drive molecular dynamics and are readily obtained using automatic differentiation. Higher-order derivatives, most notably the Hessian, describe collective motion and allow the direct prediction of experimental observables, but are considered computationally inaccessible for large systems. We suggest a solution: in physical systems, interactions decay with distance, and most MLIPs build on this locality through message passing up to a finite receptive field. This implies both sparsity of higher-order derivatives and their decay with distance. This structure can be exploited using automatic sparse differentiation (ASD). We explain how to compute the sparsity pattern for MLIP derivatives and demonstrate that, for multiple foundation MLIPs, ASD computes full Hessians of large porous materials exactly, but with modest speedups at best. The larger gains come from truncated ASD: discarding small, but nonzero, Hessian entries between distant atoms yields order-of-magnitude speedups with negligible impact on predicted observables.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-18 in cs.LG).
Categorie arXiv: Chemical Physics (physics.chem-ph) ; Materials Science (cond-mat.mtrl-sci); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.20510 — 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 2026-10-10T03:48:18.459Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-18?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).