BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20260509T224721EDT-4267eFWsru@132.216.98.100
DTSTAMP:20260510T024721Z
DESCRIPTION:Title: Efficient and Modular Implicit Differentiation.\n\nAbstr
 act: Automatic differentiation (autodiff) has revolutionized machine learn
 ing. It allows expressing complex computations by composing elementary one
 s in creative ways and removes the burden of computing their derivatives b
 y hand. More recently\, differentiation of optimization problem solutions 
 has attracted widespread attention with applications such as optimization 
 layers\, and in bi-level problems such as hyper-parameter optimization and
  meta-learning. However\, so far\, implicit differentiation remained diffi
 cult to use for practitioners\, as it often required case-by-case tedious 
 mathematical derivations and implementations. In this paper\, we propose a
  unified\, efficient and modular approach for implicit differentiation of 
 optimization problems. In our approach\, the user defines directly in Pyth
 on a function F capturing the optimality conditions of the problem to be d
 ifferentiated. Once this is done\, we leverage autodiff of F and implicit 
 differentiation to automatically differentiate the optimization problem. O
 ur approach thus combines the benefits of implicit differentiation and aut
 odiff. It is efficient as it can be added on top of any state-of-the-art s
 olver and modular as the optimality condition specification is decoupled f
 rom the implicit differentiation mechanism. We show that seemingly simple 
 principles allow to recover many exiting implicit differentiation methods 
 and create new ones easily. We demonstrate the ease of formulating and sol
 ving bi-level optimization problems using our framework. We also showcase 
 an application to the sensitivity analysis of molecular dynamics..\n\n \n
 \n \n\nZoom : https://us06web.zoom.us/j/85327310903?pwd=SlhEak53S2xrNkVYKz
 l4YUd5KzBudz09\n\nMeeting ID: 853 2731 0903 \n\nPasscode: 383854\n
DTSTART:20220207T210000Z
DTEND:20220207T220000Z
SUMMARY:Fabian Pedregosa (Google)
URL:https://www.mcgill.ca/mathstat/channels/event/fabian-pedregosa-google-3
 37389
END:VEVENT
END:VCALENDAR