BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:ML-guided Materials Discovery - Janosh Riebesell\, University of C
 alifornia\, Berkeley
DTSTART:20240205T140000Z
DTEND:20240205T143000Z
UID:TALK211135@talks.cam.ac.uk
CONTACT:Eszter Varga-Umbrich
DESCRIPTION:ML energy models have made significant leaps over the past 3 y
 ears. I will present Matbench Discovery\, a benchmark designed to measure 
 how useful ML actually is in guiding prospective materials discovery\, qua
 ntify the kind of acceleration we can expect in future discovery efforts\,
  and chart progress over time to determine which ML method performs best a
 t thermodynamic stability prediction from unrelaxed crystal structures.\nI
  will also present preliminary results which suggest that phonons from fou
 ndational force fields (esp. equivariant ones) can help us go beyond therm
 odynamic stability and predict dynamic stability across material space ver
 y cheaply and at unprecedented scale\, paving the way for future high-thro
 ughput discovery with increased fidelity. I will point out unresolved prob
 lems stemming from over-softened potential energy surfaces (PES) in these 
 foundation models and how we might address them.
LOCATION:Zoom link: https://zoom.us/j/92447982065?pwd=RkhaYkM5VTZPZ3pYSHpt
 UXlRSkppQT09
END:VEVENT
END:VCALENDAR