@elduvelle You may be happy to learn that during a PhD defense I just attended to, the student used in introduction a very nice place cell figure borrowed from... Grieves & Duvelle! 😁
=> More informations about this toot | View the thread
You can now read it here: http://doi.org/10.1111/ejn.16635
=> More informations about this toot | View the thread
https://www.cnrs.fr/fr/actualite/entre-preservation-et-mise-en-valeur-les-archives-du-cnrs
=> More informations about this toot | View the thread
@PubPeer Ah, finally, it will be published in EJN!
=> More informations about this toot | View the thread
Any of you knows about experimental setups to test sense of humor in animals?
Very serious question: send me references if you know some.
[edit: some very interesting answers below!]
=> More informations about this toot | View the thread
I still teach that unsupervised learning provides nice explanations for HPC activity.
=> More informations about this toot | View the thread
I just can't buy that.
=> More informations about this toot | View the thread
I just can't buy that.
=> More informations about this toot | View the thread
Once upon a time, it was considered that the hippocampus was learning regularities in the input flow by mean of unsupervised learning.
A type of learning adapted to that function.
And now that the hammer of backprop and Deep Learning is getting bigger and bigger, more and more models use backprop from motor output up to sensory input, thus implying that the HPC is a nail like any other, driven by supervised learning.
=> More informations about this toot | View the thread
TICS refused to take in consideration the publication of a commentary of a paper they published 5 months ago, because it was not submitted "within 3 or 4 months" (appreciate the precision). Even when we pointed out that there was something wrong in the paper.
Well, it will end up on pubpeer... @PubPeer
=> More informations about this toot | View the thread
About today's shitstorm:
a dozen of years ago, my office mates and myself were regularly using the "Schmiduber-Grossberg scale, numbered from 1 to 10".
I let the reader imagine what academic phenomenon it was measuring.
=> More informations about this toot | View the thread
La signalétique est posée !
=> View attached media | View attached media
=> More informations about this toot | View the thread
Pour les curieux.ses, je serai là :
=> More informations about this toot | View the thread
Présentations :
=> More informations about this toot | View the thread
...
=> More informations about this toot | View the thread
Démonstrations :
...
=> More informations about this toot | View the thread
Très beau programme à l'ISIR cette année pour la fête de la science !
Venez nous voir : campus Pierre & Marie Curie à Paris (Métro Jussieu).
Sur le site, il y aura plein de labos faisant des présentations sur le parvis sous des barnums (difficile de les rater).
Les activités de l'ISIR sont pour beaucoup dans le laboratoire, du côté des tours 55 et 66. Ce sera fléché (normalement).
=> More informations about this toot | View the thread
This result probably tells us that my community on Mastodon is more neuroscience that AI oriented ! ^_^
=> More informations about this toot | View the thread
Hopfiel nets: physics, AI or computational neuroscience?
^_^
=> More informations about this toot | View the thread
I have a question, following a discussion we had with colleagues at lunch: for you, is it a problem that the backpropagation algorithm does not have a clearly identified biological mechanisms?
I have the feeling that this problem was considered important 20 years ago, but I don't see so many people mentionning it nowadays. I am not sure a definitive answer to this problem has been given (if so, tell me about it!), and, well, deep learning is just backprop, backprop everywhere...
=> More informations about this toot | View the thread
=> This profile with reblog | Go to BenoitGirard@sciences.re account This content has been proxied by September (ba2dc).Proxy Information
text/gemini