BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Neural networks and interfaces: theoretical considerations and pra
 ctical solutions - Samuel Niblett\, Lawrence Berkeley National Laboratory
DTSTART:20211206T160000Z
DTEND:20211206T163000Z
UID:TALK162346@talks.cam.ac.uk
CONTACT:Dr Christoph Schran
DESCRIPTION:Reactive machine-learned potentials have recently emerged as o
 ne of the most promising computational tools for materials chemistry\, all
 owing us to apply statistical mechanical methods to more realistic and com
 plex problems than ever before. Many of these problems require training mo
 dels of exotic environments and non-equilibrium boundary conditions\, wher
 e existing model architectures and training protocols may not be suitable.
  These limitations are only now beginning to be explored.\n\nIn this talk\
 , I will show that the inherent localisation of most symmetry functions pr
 events standard machine-learned models from describing liquid-vapour inter
 faces correctly\, since many properties of these interfaces are sensitive 
 to unbalanced long-range forces such as electrostatic interactions. Using 
 the perspective of local molecular field (LMF) theory due to Weeks and cow
 orkers\, I will show how neural network potentials attempt to represent th
 ese long-ranged forces and where they fail. The LMF approach suggests a si
 mple modification to the neural network method that dramatically improves 
 the description of polarisation-dependent properties. I will also discuss 
 practical details of our neural network protocols\, including refinements 
 to the standard loss functions and our approach to data generation\, both 
 of which improve training efficiency and model generalisation.
LOCATION:BMS Lecture Theatre\, Department of Chemistry
END:VEVENT
END:VCALENDAR