BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Computational Neuroscience Journal Club - Yashar Ahmadian and Puri
 a Radmard
DTSTART:20230620T130000Z
DTEND:20230620T150000Z
UID:TALK202678@talks.cam.ac.uk
CONTACT:Luke Johnston
DESCRIPTION:Please join us for our fortnightly Computational Neuroscience 
 journal club on Tuesday 20th June at 2pm UK time in the CBL seminar room\,
  or online on zoom. The title is ‘Geometry of neural population response
 s’\, presented by Yashar Ahmadian and Puria Radmard.\n\nZoom information
 : https://eng-cam.zoom.us/j/84204498431?pwd=Um1oU284b1YxWThObGw4ZU9XZitWdz
 09 Meeting ID: 842 0449 8431 Passcode: 684140\n\nSummary:\n\nEfficient cod
 ing theories have long hypothesized that neural representations should min
 imize redundancy and  correlations between neurons’ responses (at least 
 when noise does not dominate). Low correlations are equivalent to a high d
 imensionality of the coding subspace (the subspace of the neural response 
 space spanned by signals). On the other hand\, in many experiments stimulu
 s representations and task relevant variables seem to furnish a low-dimens
 ional representation. However\, this might simply reflect the artificially
  low complexity of typical experiments (e.g. due to a small or low-dimensi
 onal set of stimuli\, or simple tasks).\nIn the first paper we will presen
 t\, Stringer et al. (2019) present a large ensemble of natural images to m
 ice and record simultaneously from thousands of neurons in their V1. They 
 find a high-dimensional representation of natural images\, with a signal c
 ovariance spectrum that drops as a scale-invariant power-law 1/n. Moreover
  they mathematically derive a bound on the decay of this spectrum for smoo
 th population codes on d-dimensional stimulus manifolds\, and conclude tha
 t the mouse V1 operates close to this limit for large d.\nWang and colleag
 ues (2023) extend this analysis to whole brain activity in zebrafish. In t
 his case\, they find that dimensionality remains well above the critical l
 imit for spontaneous and behaviour related activity\, both for the whole b
 rain and for random subsets of neurons. They further use a Euclidean Rando
 m Matrix model to provide a functional embedding for the neural population
 \, showing anatomical clustering in this space.\n\nReferences:\n\n[1] Stri
 nger et al.\, 2019: High-dimensional geometry of population responses in v
 isual cortex (https://www.nature.com/articles/s41586-019-1346-5)\n\n[2] Wa
 ng et al.\, 2023: The scale-invariant covariance spectrum of brain-wide ac
 tivity (https://www.biorxiv.org/content/10.1101/2023.02.23.529673v1)
LOCATION:In Person (CBL Seminar Room) and Online on Zoom
END:VEVENT
END:VCALENDAR