BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20250712T004817EDT-8946tHVZNC@132.216.98.100
DTSTAMP:20250712T044817Z
DESCRIPTION:Title: Dynamics of high-dimensional recurrent networks: how cha
 os shapes computation in biological and artificial neural networks\n\nAbst
 ract: Networks of neurons -either biological or artificial- are called rec
 urrent if their connections are distributed and contain feedback loops. Su
 ch networks can perform remarkably complex computations\, as evidenced by 
 their ubiquity throughout the brain and ever-increasing use in machine lea
 rning. They are\, however\, notoriously hard to control and their dynamics
  are generally poorly understood\, especially in the presence of external 
 forcing. This is because recurrent networks are typically chaotic systems\
 , meaning they have rich and sensitive dynamics leading to variable respon
 ses to inputs. How does this chaos manifest in the neural code of the brai
 n? How might we tame sensitivity to exploit complexity when training artif
 icial recurrent networks for machine learning? Understanding how the dynam
 ics of large driven networks shape their capacity to encode and process in
 formation presents a sizeable challenge. In this talk\, I will discuss the
  use of Random Dynamical Systems Theory as a framework to study informatio
 n processing in high-dimensional\, signal-driven networks. I will present 
 an overview of recent results linking chaotic attractors to entropy produc
 tion\, dimensionality and input discrimination of dynamical observables. I
  will outline insights this theory provides on how cortex performs complex
  computations using sparsely connected inhibitory and excitatory neurons\,
  as well as implications for gradient-based optimization methods for artif
 icial networks.\n
DTSTART:20180409T200000Z
DTEND:20180409T200000Z
LOCATION:Room 920\, Burnside Hall\, CA\, QC\, Montreal\, H3A 0B9\, 805 rue 
 Sherbrooke Ouest
SUMMARY:Guillaume Lajoie (Université de Montréal)
URL:/mathstat/channels/event/guillaume-lajoie-universi
 te-de-montreal-286332
END:VEVENT
END:VCALENDAR