An SO(3)-equivariant reciprocal-space neural potential for long-range interactions

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Zhang, Lingfeng, Cui, Taoyong, Zhou, Dongzhan, Bai, Lei, Zhang, Sufei, Rossi, Luca, Su, Mao, Ouyang, Wanli, Heng, Pheng-Ann
Formato: Preprint
Publicado: 2026
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866914409754918912
author Zhang, Lingfeng
Cui, Taoyong
Zhou, Dongzhan
Bai, Lei
Zhang, Sufei
Rossi, Luca
Su, Mao
Ouyang, Wanli
Heng, Pheng-Ann
author_facet Zhang, Lingfeng
Cui, Taoyong
Zhou, Dongzhan
Bai, Lei
Zhang, Sufei
Rossi, Luca
Su, Mao
Ouyang, Wanli
Heng, Pheng-Ann
contents Long-range electrostatic and polarization interactions play a central role in molecular and condensed-phase systems, yet remain fundamentally incompatible with locality-based machine-learning interatomic potentials. Although modern SO(3)-equivariant neural potentials achieve high accuracy for short-range chemistry, they cannot represent the anisotropic, slowly decaying multipolar correlations governing realistic materials, while existing long-range extensions either break SO(3) equivariance or fail to maintain energy-force consistency. Here we introduce EquiEwald, a unified neural interatomic potential that embeds an Ewald-inspired reciprocal-space formulation within an irreducible SO(3)-equivariant framework. By performing equivariant message passing in reciprocal space through learned equivariant k-space filters and an equivariant inverse transform, EquiEwald captures anisotropic, tensorial long-range correlations without sacrificing physical consistency. Across periodic and aperiodic benchmarks, EquiEwald captures long-range electrostatic behavior consistent with ab initio reference data and consistently improves energy and force accuracy, data efficiency, and long-range extrapolation. These results establish EquiEwald as a physically principled paradigm for long-range-capable machine-learning interatomic potentials.
format Preprint
id arxiv_https___arxiv_org_abs_2603_18389
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle An SO(3)-equivariant reciprocal-space neural potential for long-range interactions
Zhang, Lingfeng
Cui, Taoyong
Zhou, Dongzhan
Bai, Lei
Zhang, Sufei
Rossi, Luca
Su, Mao
Ouyang, Wanli
Heng, Pheng-Ann
Chemical Physics
Artificial Intelligence
Long-range electrostatic and polarization interactions play a central role in molecular and condensed-phase systems, yet remain fundamentally incompatible with locality-based machine-learning interatomic potentials. Although modern SO(3)-equivariant neural potentials achieve high accuracy for short-range chemistry, they cannot represent the anisotropic, slowly decaying multipolar correlations governing realistic materials, while existing long-range extensions either break SO(3) equivariance or fail to maintain energy-force consistency. Here we introduce EquiEwald, a unified neural interatomic potential that embeds an Ewald-inspired reciprocal-space formulation within an irreducible SO(3)-equivariant framework. By performing equivariant message passing in reciprocal space through learned equivariant k-space filters and an equivariant inverse transform, EquiEwald captures anisotropic, tensorial long-range correlations without sacrificing physical consistency. Across periodic and aperiodic benchmarks, EquiEwald captures long-range electrostatic behavior consistent with ab initio reference data and consistently improves energy and force accuracy, data efficiency, and long-range extrapolation. These results establish EquiEwald as a physically principled paradigm for long-range-capable machine-learning interatomic potentials.
title An SO(3)-equivariant reciprocal-space neural potential for long-range interactions
topic Chemical Physics
Artificial Intelligence
url https://arxiv.org/abs/2603.18389