Metraipolis › archive › dc_35d4dab49367
BranchIP: Learning Adaptive Equivariant Computation for Interatomic Potentials — arXiv:2610.02013
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.02013
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T12:07:29.777Z
- SHA-256 of the archived text
a5bf7315f9e3849c1ee74b5f6a1357faf305d89b093854b93b8ffeef4dfea249- Stable URL of this page
https://metraipolis.com/documento/dc_35d4dab49367
How to cite
BranchIP: Learning Adaptive Equivariant Computation for Interatomic Potentials — arXiv:2610.02013 — arXiv (en), https://arxiv.org/abs/2610.02013 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_35d4dab49367 (Metraipolis archive dc_35d4dab49367, sha256 a5bf7315f9e3849c1ee74b5f6a1357faf305d89b093854b93b8ffeef4dfea249)
BibTeX
@misc{dc35d4dab49367,
title = {BranchIP: Learning Adaptive Equivariant Computation for Interatomic Potentials — arXiv:2610.02013},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.02013},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_35d4dab49367; sha256 a5bf7315f9e3849c1ee74b5f6a1357faf305d89b093854b93b8ffeef4dfea249},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.02013. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Equivariant machine learning interatomic potentials (MLIPs) have revolutionized atomistic modeling, but accurate treatment of complex materials and molecular systems demands expensive models. This limits simulation length- and time-scales, with tensor products a key computational bottleneck. The recent emergence of foundation-scale MLIPs further exacerbates this challenge. We present Branch Interatomic Potential (BranchIP), a single-model framework for learned adaptive tensor product computation, trained with a novel distillation loss. In our experiments on two systems of physical interest, a heterogeneous catalysis system and a proton-conducting solid acid electrolyte, BranchIP accelerates MLIPs across model sizes by up to $2.4\times$ while reducing memory usage by up to $2.6\times$. This is achieved while maintaining physical fidelity. Furthermore, the learned adaptive computation provides model interpretability by revealing which interactions demand deeper computation and showing how computational depth relates to chemical complexity and dynamics. Tema degli esploratori: fisica e astronomia (retro-catalogo arXiv: annunci del 2026-10-02 in physics.comp-ph). Categorie arXiv: Machine Learning (cs.LG) ; Applied Physics (physics.app-ph); Chemical Physics (physics.chem-ph); Computational Physics (physics.comp-ph). Pagina del lavoro: https://arxiv.org/abs/2610.02013 — 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 2026-10-07T11:38:27.201Z dalla pagina https://arxiv.org/catchup/physics.comp-ph/2026-10-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).