Condensed Matter > Materials Science
[Submitted on 2 Aug 2022 (v1), revised 18 Jul 2023 (this version, v2), latest version 26 Feb 2024 (v3)]
Title:Permutation-adapted complete and independent basis for atomic cluster expansion descriptors
View PDFAbstract:For the purposes of interatomic potentials, atomic cluster expansion (ACE) methods provide a generalizable and systematic way to describe atomic environments of arbitrary body order. As an extension of fixed-lattice cluster expansions, it may be used to model properties for arbitrary collections of atoms with or without translational symmetry. This makes them excellent choices for machine-learned interatomic potentials. Extending cluster expansions requires construction of a rotation and permutation invariant (RPI) cluster basis. When these functions are symmetrized, this basis becomes over-complete. Current approaches for ACE methods involve a numerical step for this reason. This prevents a full analytical definition of an ACE basis and ACE descriptors, and can be susceptible to numerical instabilities. In this work, we propose a method that adapts to this permutation invariance to obtain a rotation and permutation invariant (PA-RPI) basis that is complete and independent. Through the presentation of this method, we demonstrate how and why there are linearly dependent basis functions through some explicit proofs. The proofs as well as our method leverage the permutation symmetries as well as recursion relationships of the generalized Wigner symbols. These properties of the generalized Wigner symbols are used to derive relationships between RPI functions. Following derivations of these properties, the procedure for the construction of the PA-RPI basis is presented. Among some specific cases of RPI basis construction, we provide extensive tabulations of PA-RPI basis function counts. We also show that in practical applications for interatomic potentials, high degree descriptors help achieve high-accuracy models. High degree, high body-order functions can now be selected based on analytical recursion relationships and used in ACE models rather than relying on numerical SVD.
Submission history
From: James Goff [view email][v1] Tue, 2 Aug 2022 21:13:44 UTC (1,064 KB)
[v2] Tue, 18 Jul 2023 01:28:02 UTC (2,223 KB)
[v3] Mon, 26 Feb 2024 17:37:46 UTC (1,585 KB)
Current browse context:
cond-mat.mtrl-sci
Change to browse by:
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
IArxiv Recommender
(What is IArxiv?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.