BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Computational Neuroscience Journal Club - Jake Stroud and David Li
 u
DTSTART:20221025T130000Z
DTEND:20221025T150000Z
UID:TALK185879@talks.cam.ac.uk
CONTACT:Luke Johnston
DESCRIPTION:Please join us for our fortnightly journal club online via zoo
 m where two presenters will jointly present a topic together. The next top
 ic is ‘Analysis of RNNs’ presented by Jake Stroud and David Liu.\n\nZo
 om information: https://us02web.zoom.us/j/84958321096?pwd=dFpsYnpJYWVNeHlJ
 bEFKbW1OTzFiQT09 Meeting ID: 849 5832 1096 Passcode: 506576\n\nSummary: Re
 current neural networks (RNNs) have been successfully used to aid our unde
 rstanding many different brain regions from visual cortex\, to the basal g
 anglia\, hippocampus\, and frontal cortices. Historically\, these network 
 models (and in particular the connections between neurons in the network) 
 were hand crafted so that the model displayed the desired behaviour (e.g. 
 Hopfield networks). Recently\, with the advent of new machine learning tec
 hniques\, it has become increasingly more common for researchers to instea
 d train RNNs on a given task and then subsequently reverse-engineer the ne
 twork to understand how it solved the task. In this journal club\, we will
  present the techniques that allow us to uncover the mechanisms that train
 ed RNNs use to solve tasks. In particular\, we will show why fixed points 
 analyses\, and local linearizations around fixed points\, are central to u
 nderstanding the dynamics of RNNs. We will then cover a recent pre-print t
 hat uses these approaches to understand how a single RNN performs multiple
  different tasks.\n\nKey papers:\n\nOpening the black box: Low-dimensional
  dynamics in high-dimensional recurrent neural networks\, David Sussillo a
 nd Omri Barak\, Neural Computation\, 2013.\n\nNeural circuits as computati
 onal dynamical systems\, David Sussillo\, Current Opinion in Neurobiology\
 , 2014.\n\nFlexible multitask computation in recurrent networks utilizes s
 hared dynamical motifs\, Laura Driscoll\, Krishna Shenoy\, David Sussillo\
 , BioRxiv\, 2022\n
LOCATION:Online on Zoom
END:VEVENT
END:VCALENDAR