It's a bit like comparing the number of atoms in a glass of mercury with the total number of cells in the grapes that went into making your glass of wine It makes no sense whatsoever
-
-
Show this threadThanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
What is the estimated overhead for a Feedforward RNN to emulate a real biological neuron's evolution in time?
-
This would only be relevant if feedforward RNNs were actually used to emulate biological neurons. Otherwise the only common ground is that you have large natural integers on both sides
- Show replies
New conversation -
-
-
This Tweet is unavailable.
-
I struggle to add & multiply numbers, my own brain is definitely under 1 flops
- Show replies
-
-
-
This thinking is also in the Deep Learning book, which goes from explaining ANN units as approximations of biological neurons to how humans have roughly 80bn of those.. Trend may increase as more students open their eyes to the field through studying that book.
-
Goodfellow deals quite a lot with brain comparison in the introduction section of the book. For example he says: ”one should not view deep learning as an attempt to simulate the brain.” and ”We know that actual neurons compute very different functions than modern ReLUs” (pg. 16).
- Show replies
New conversation -
-
-
So you say that comparisons made by
@goodfellow_ian et al. in 'Deep Learning' Page 22/23 (https://www.deeplearningbook.org/ ) are braindead?Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
This Tweet is unavailable.
-
Or do they see at 224x224x3?!
End of conversation
-
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.