BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Syntax generation through synfire propagation
DTSTART:20101217T121500
DTSTAMP:20260407T020927Z
UID:3cc72755388f7873b7da267ea0663eb2d7ead9094bd437e2120cfeea
CATEGORIES:Conferences - Seminars
DESCRIPTION:Abigail MORRISON\nAdult Bengalese finches generate a variable 
 song that obeys a distinct and individual syntax. The syntax is gradually 
 lost over a period of days after deafening and is recovered when hearing i
 s restored. In the first part of this talk I will present a spiking neuron
 al network model of the song syntax generation and its loss\, based on the
  assumption that the syntax is stored in reafferent connections from the a
 uditory to the motor control area. Propagating synfire activity in the HVC
  (high vocal center) codes for individual syllables of the song and primin
 g signals from the auditory network reduce the competition between syllabl
 es to allow only those transitions that are permitted by the syntax. Both 
 imprinting of song syntax within HVC and the interaction of the reafferent
  signal with an efference copy of the motor command are sufficient to expl
 ain the gradual loss of syntax in the absence of auditory feedback. This s
 tudy illustrates how sequential compositionality following a defined synta
 x can be realized in networks of spiking neurons. \nIn the second part of 
 this talk I will consider how the synfire chains assumed in the first part
  could develop. It has long been though that spike-timing dependent plasti
 city (STDP) provides an answer to the question of how the brain can develo
 p functional structure in response to repeated stimuli. However\, convinci
 ng demonstrations of this capacity in large\, initially random networks ha
 ve not been forthcoming\; such demonstrations as there are typically rely 
 on constraining the problem artificially. I will present a theoretical ana
 lysis based on a mean field approach of the development of feed-forward st
 ructure in random networks. An unstable fixed point in the recruitment dyn
 amics prevents the stable propagation of structure in recurrent networks w
 ith weight-dependent STDP. We demonstrate that the key predictions of the 
 theory hold in large-scale simulations. The theory provides insight into t
 he reasons why such development does not take place in unconstrained syste
 ms and enables us to identify candidate biologically motivated adaptations
  to the balanced random network model that might resolve the issue.
LOCATION:BC 01 https://plan.epfl.ch/?room==BC%2001
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR