BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Four Generations of Neural Network Potentials - Jörg Behler\, Geo
 rg-August-Universität Göttingen
DTSTART:20220214T140000Z
DTEND:20220214T143000Z
UID:TALK167279@talks.cam.ac.uk
CONTACT:Dr Christoph Schran
DESCRIPTION:A lot of progress has been made in recent years in the develop
 ment of machine learning potentials (MLP) for atomistic simulations [1]. N
 eural network potentials\n(NNPs)\, which have been introduced more than tw
 o decades ago [2]\, are an important class of MLPs. While the first genera
 tion of NNPs has been restricted to small\nmolecules with only a few degre
 es of freedom\, the second generation extended the applicability of MLPs t
 o high-dimensional systems containing thousands of atoms by\nconstructing 
 the total energy as a sum of environment-dependent atomic energies [3]. Lo
 ng-range electrostatic interactions can be included in third-generation NN
 Ps\nemploying environment-dependent charges [4]\, but only recently limita
 tions of this locality approximation could be overcome by the introduction
  of fourth-generation NNPs [5\,6]\, which are able to describe non-local c
 harge transfer using a global charge\nequilibration step. In this talk an 
 overview about the evolution of high-dimensional neural network potentials
  will be given along with typical applications in large-scale atomistic si
 mulations.\n\n[1] J. Behler\, J. Chem. Phys. 145 (2016) 170901.\n\n[2] T. 
 B. Blank\, S. D. Brown\, A. W. Calhoun\, and D. J. Doren\, J. Chem. Phys. 
 103 (1995) 4129.\n\n[3] J. Behler and M. Parrinello\, Phys. Rev. Lett. 98 
 (2007) 146401.\n\n[4] N. Artrith\, T. Morawietz\, J. Behler\, Phys. Rev. B
  83 (2011) 153101.\n\n[5] S. A. Ghasemi\, A. Hofstetter\, S. Saha and S. G
 oedecker\, Phys. Rev. B 92 (2015) 045131.\n\n[6] T. W. Ko\, J. A. Finkler\
 , S. Goedecker\, J. Behler\, Nature Comm. 12 (2021) 398.
LOCATION:Venue to be confirmed
END:VEVENT
END:VCALENDAR