BEGIN:VCALENDAR
VERSION:2.0
CALSCALE:GREGORIAN
PRODID:UW-Madison-Physics-Events
BEGIN:VEVENT
SEQUENCE:0
UID:UW-Physics-Event-6598
DTSTART:20210908T160000Z
DTEND:20210908T171500Z
DTSTAMP:20260414T192103Z
LAST-MODIFIED:20210903T201759Z
LOCATION:Online Seminar: Please sign up for our mailing list at www.ph
 ysicsmeetsml.org for zoom link
SUMMARY:JAX MD: A Framework for Differentiable Atomistic Physics\, Phy
 sics ∩ ML Seminar\, Sam Schoenholz\, Google Brain
DESCRIPTION:I will talk about JAX MD\, a software package for performi
 ng differentiable physics simulations with a focus on molecular dynami
 cs. JAX MD includes a number of physics simulation environments\, as w
 ell as interaction potentials and neural networks that can be integrat
 ed into these environments without writing any additional code. Since 
 the simulations themselves are differentiable functions\, entire traje
 ctories can be differentiated to perform meta-optimization. These feat
 ures are built on primitive operations\, such as spatial partitioning\
 , that allow simulations to scale to hundreds-of-thousands of particle
 s on a single GPU. My talk will include an introduction to the JAX sof
 tware package www.github.com/google/jax. If you are interested in tryi
 ng out JAX MD\, it is available at github.com/google/jax-md.
URL:https://www.physics.wisc.edu/events/?id=6598
END:VEVENT
END:VCALENDAR