The design space of E(3)-equivariant atom-centred interatomic potentials

Molecular dynamics simulation is an important tool in computational materials science and chemistry, and in the past decade it has been revolutionized by machine learning. This rapid progress in machine learning interatomic potentials has produced a number of new architectures in just the past few years. Particularly notable among these are the atomic cluster expansion, which unified many of the earlier ideas around atom-density-based descriptors, and Neural Equivariant Interatomic Potentials (NequIP), a message-passing neural network with equivariant features that exhibited state-of-the-art accuracy at the time. Here we construct a mathematical framework that unifies these models: atomic cluster expansion is…

• NS
  National Science Foundation

  Awards: 2011754, DMR-2011754
• UD
  U.S. Department of Energy

  Awards: DE-SC0022199, DE-SC0021110, SC0021110, DE-SC0012573
• A
  AstraZeneca
• HU
  Harvard University

  Award: DMR-2011754
• DE
  Dell EMC
• LT
  Leverhulme Trust

  Award: RPG-2017-191
• OO
  Office of Science

  Awards: DE-SC0022199, DE-SC0021110, DE-SC0012573
• MR
  Materials Research Science and Engineering Center, Harvard University

  Awards: DMR-2011754, 2011754
• NS
  Natural Sciences and Engineering Research Council of Canada

  Award: IDGR019381
• EA
  Engineering and Physical Sciences Research Council

  Awards: EP/T022159/, EP/T022159/1
• SA
  Science and Technology Facilities Council
• DO
  Division of Materials Research

  Awards: DMR-2011754, 2011754
• BE
  Basic Energy Sciences

  Awards: -SC0012573, DE-SC0021110, DE-SC0012573, DE-SC0022199
• AS
  Advanced Scientific Computing Research

  Award: DE-SC0021110
• FD
  FAS Division of Science, Harvard University