Gave a talk on why I think backpropagation is a good model for aspects of the brain for #USS2020. Watch the recording here: https://youtu.be/bZ6bgo8j-Hw?t=402 …
-
-
Replying to @KordingLab
In your opinion, which specific electrophysiological setting would be the better one to look for the backpropagation mechanism in the brain? LFPs, Calcium imaging or Parallel spike trains/single unit ?
1 reply 0 retweets 0 likes -
Replying to @NunezKant @KordingLab
Which would be the criteria to select a specific microcircuit in the brain or this property should be generalized across the brain ? I mean, BProp in your opinion should be a neuron property or a circuit property ?
1 reply 0 retweets 1 like -
Replying to @NunezKant
I don't really know.
1 reply 0 retweets 2 likes -
Replying to @KordingLab @NunezKant
Maybe cerebellum https://elifesciences.org/articles/31599 & basal ganglia do node perturbation https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3221789/ … Hippo->short term pattern completion (no BP) Ctx->long term pattern completion (BP for “filling in”), & other (BP w/ area-specific costs)->dendritic mechanisms in circuit context
1 reply 0 retweets 2 likes -
Could be 96% wrong. But striving for coherence is good. As long as other pictures that also strive for coherence, or expose incoherence, are also celebrated.
1 reply 0 retweets 1 like -
In science I think we need more ability for ideas to get to scale. Including DL inspired pictures of brain function. Just not to a scale where they entirely take over. 5-10 big pictures rather than 5000 tiny pictures or 1 monster picture.
2 replies 0 retweets 7 likes
@sir_deenicus: in response to your gravitational field comment
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.