BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Department of Theoretical Physics - ECPv6.5.1.5//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-ORIGINAL-URL:https://web-f1.ijs.si
X-WR-CALDESC:Events for Department of Theoretical Physics
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:Europe/Ljubljana
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:20210328T010000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:20211031T010000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=Europe/Ljubljana:20211109T130000
DTEND;TZID=Europe/Ljubljana:20211109T140000
DTSTAMP:20260427T233639
CREATED:20210924T082053Z
LAST-MODIFIED:20211103T092853Z
UID:4579-1636462800-1636466400@web-f1.ijs.si
SUMMARY:[Biophysics seminar] Julija Zavadlav: "Neural network potentials trained top-down"
DESCRIPTION:Molecular modeling has become a cornerstone of many disciplines\, including soft matter physics. However\, the quality of predictions critically depends on the employed model that defines particle interactions. A class of models with tremendous success in recent years are neural network (NN) potentials due to their flexibility and capacity of learning many-body interactions. Traditionally\, these models are trained bottom-up on quantum mechanical data. Top-down approaches that learn NN potentials directly from experimental data have received less attention\, typically facing numerical and computational challenges when backpropagating through molecular dynamics (MD) simulations. We recently developed the Differentiable Trajectory Reweighting (DiffTRe) method\, which bypasses differentiation through the MD simulation for time-independent observables. Leveraging thermodynamic perturbation theory\, we avoid exploding gradients and achieve around 2 orders of magnitude speed-up in gradient computation for top-down learning. The effectiveness of DiffTRe is showcased on atomistic and coarse-grained models using diverse experimental observables including thermodynamic\, structural\, and mechanical properties. Our approach opens the way to high fidelity molecular models\, particularly when bottom-up data is unavailable or insufficiently accurate.\n\n\nThe meeting will take place online via Zoom: \nMeeting ID: 833 0750 1179 \nPasscode: bio
URL:https://web-f1.ijs.si/event/biophysics-seminar-julija-zavadlav-nn-potentials/
LOCATION:Zoom
END:VEVENT
END:VCALENDAR