BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VTIMEZONE
TZID:Europe/Vienna
BEGIN:DAYLIGHT
DTSTART:20240331T030000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20241027T020000
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260405T162347Z
UID:66a7436d57c41722650113@ist.ac.at
DTSTART:20240730T130000
DTEND:20240730T140000
DESCRIPTION:Speaker: Flavio Martinelli\nhosted by Tim Vogels\nAbstract: Can
  we identify the weights of an artificial neural network by probing its in
 put-output mapping? At first glance\, this problem seems to have too many 
 solutions because of various symmetries. Yet\, we show that the incoming w
 eight vector of each neuron is identifiable up to sign or scaling\, depend
 ing on the activation function. Our novel approach 'Expand-and-Cluster can
  identify layer sizes and weights of a target network for all commonly use
 d activation functions.In my talk\, I will give some intuitions on the dyn
 amics of training for a very specific learning setup: a network that regre
 sses the output of another network (teacher-student setup). Then\, I will 
 show you how we avoid local minima of gradient descent dynamics and why id
 entification of weights is possible. To conclude\, I will present a practi
 cal algorithm to perform network identification.
LOCATION:Office Bldg West / Ground floor / Heinzel Seminar Room (I21.EG.101
 )\, ISTA
ORGANIZER:jdeanton@ist.ac.at
SUMMARY:Flavio Martinelli: From Loss Landscape Geometry to Weight Recovery 
 in Neural Networks
URL:https://talks-calendar.ista.ac.at/events/5118
END:VEVENT
END:VCALENDAR