BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20251111T143229EST-5655xu7CPt@132.216.98.100
DTSTAMP:20251111T193229Z
DESCRIPTION:Supported by the generosity of the Killam Trusts\, the MNI's Ki
 llam Seminar Series invites outstanding guest speakers whose research is o
 f interest to the scientific community at the MNI and ɬÀï·¬.\n
 \n\nTo attend in person\, register (here)\n\nIn-person talk only. No virtu
 al option. \n\n\nCarsen Stringer\, PhD\n\nGroup leader\, HHMI Janelia Rese
 arch Campus\, Virginia\, USA\n\nHost: Stuart Trenholm\n\nAbstract: Represe
 ntation learning in neural networks may be implemented with supervised or 
 unsupervised algorithms\, distinguished by the presence or absence of rewa
 rd feedback. Both types of learning are highly effective in artificial neu
 ral networks. In biological systems\, task learning has been shown to chan
 ge sensory neural representations\, but it is not known if these changes a
 re due to supervised or unsupervised learning. Here we recorded population
 s of ~70\,000 neurons simultaneously from primary visual cortex (V1) and h
 igher visual areas (HVA)\, while mice learned multiple tasks as well as du
 ring unrewarded exposure to the same stimuli. We find that neural changes 
 due to task learning were concentrated in medial and anterior HVAs. The ch
 anges in medial HVAs were also found in mice that did not learn a task\, w
 hile changes in anterior HVAs were not. Anterior HVAs represented a rampin
 g reward anticipation signal which was abolished by the delivery of reward
 \, consistent with the involvement of this area in supervised learning. Ac
 ross different tasks\, neural changes in all areas including V1 were consi
 stent with a pattern of generalizing to new stimuli according to the rules
  of the respective task\, even when the task was not explicitly instructed
 . Thus\, most changes in neural representations in visual areas are due to
  unsupervised learning and these changes may support behavioral generaliza
 tion in ecological scenarios where rewards are rare.\n
DTSTART:20230926T200000Z
DTEND:20230926T210000Z
LOCATION:de Grandpre Communications Centre\, Montreal Neurological Institut
 e\, CA\, QC\, Montreal\, H3A 2B4\, 3801 rue University
SUMMARY:Killam Seminar Series: Unsupervised Pretraining of Neural Represent
 ations for Task Learning
URL:/neuro/channels/event/killam-seminar-series-unsupe
 rvised-pretraining-neural-representations-task-learning-349316
END:VEVENT
END:VCALENDAR