BEGIN:VCALENDAR
VERSION:2.0
METHOD:PUBLISH
PRODID:-//Missouri State University/Calendar of Events//EN
CALSCALE:GREGORIAN
X-WR-TIMEZONE:America/Chicago
BEGIN:VTIMEZONE
TZID:America/Chicago
BEGIN:DAYLIGHT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
DTSTART:20070311T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
TZNAME:CDT
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
DTSTART:20071104T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
TZNAME:CST
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
UID:e14f6c3f-e023-4c32-842a-4939d935913b.224214@calendar.missouristate.edu
CREATED:20230124T145435Z
LAST-MODIFIED:20230124T145435Z
LOCATION:
SUMMARY:PAMS Seminar: "Equivariant Interatomic Potentials" by Simon Batzner
DESCRIPTION:Simon BatznerApplied MathematicsHarvard University\n\n\nAbstra
 ct:Symmetry plays a central role in the representation of materials for t
 he purpose of Machine Learning. In particular\, all sensible representati
 ons must obey the symmetries of 3D space: translation\, rotation\, and in
 version\, in addition to permutation symmetry with respect to the labelin
 g of atoms. Traditionally\, representations have been constructed to poss
 ess invariance with respect to the above transformations. In this talk\, 
 I will discuss our efforts to generalize invariance to the broader class 
 of equivariant representations and demonstrate how this leads to a large 
 increase in generalization accuracy and sample-efficiency of the learned 
 models. The talk will then discuss the recently introduced Neural Equivar
 iant Interatomic Potential (NequIP) and Allegro potentials\, two E(3)-equ
 ivariant Interatomic Potential that exhibit unprecedented accuracy and sa
 mple efficiency and outperform invariant potentials with up to 1000x fewe
 r reference data. I will discuss applications to a diverse set of materia
 ls systems\, including Li diffusion\, amorphous structures\, heterogeneou
 s catalysis\, and water.\n\n\n\n\n\nThis is a Zoom seminar.
X-ALT-DESC;FMTTYPE=text/html:&lt;html&gt;&lt;head&gt;&lt;title&gt;&lt;/title&gt;&lt;/head&gt;&lt;body&gt;&lt;p&gt;&lt;s
 trong&gt;Simon Batzner&lt;/strong&gt;&lt;br&gt;&lt;strong&gt;Applied Mathematics&lt;br&gt;Harvard Un
 iversity&lt;/strong&gt;&lt;/p&gt;\n&lt;p&gt;Abstract:&lt;br&gt;Symmetry plays a central role in t
 he representation of materials for the purpose of Machine Learning. In pa
 rticular\, all sensible representations must obey the symmetries of 3D sp
 ace: translation\, rotation\, and inversion\, in addition to permutation 
 symmetry with respect to the labeling of atoms. Traditionally\, represent
 ations have been constructed to possess invariance with respect to the ab
 ove transformations. In this talk\, I will discuss our efforts to general
 ize invariance to the broader class of equivariant representations and de
 monstrate how this leads to a large increase in generalization accuracy a
 nd sample-efficiency of the learned models. The talk will then discuss th
 e recently introduced Neural Equivariant Interatomic Potential (NequIP) a
 nd Allegro potentials\, two E(3)-equivariant Interatomic Potential that e
 xhibit unprecedented accuracy and sample efficiency and outperform invari
 ant potentials with up to 1000x fewer reference data. I will discuss appl
 ications to a diverse set of materials systems\, including Li diffusion\,
  amorphous structures\, heterogeneous catalysis\, and water.&lt;/p&gt;\n&lt;p&gt;&lt;/p&gt;
 \n&lt;p&gt;This is a Zoom seminar.&lt;/p&gt;&lt;/body&gt;&lt;/html&gt;
DTSTART;TZID=America/Chicago:20230202T160000
DTEND;TZID=America/Chicago:20230202T170000
SEQUENCE:0
URL:https://physics.missouristate.edu/seminars.htm
CATEGORIES:Public,Alumni,Current Students,Faculty,Future Students,Staff
END:VEVENT
END:VCALENDAR