A trend that's incredibly misleading and frankly braindead: comparing the number of neurons in animals (C. elegans, honeybee, marmoset...) with the number of "neurons" in an ANN (which is the sum of the dimensionalities of the intermediate spaces you use for affine projections)
-
-
I can only think of one word at a time
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Try training a deep net to solve simple arithmetic tasks (like 21 * 13 = ?) presented as images. Even though you use up GFLOPS training or evaluating it, it will not be able to consistently perform a single FLOP on that level of abstraction.
- End of conversation
New 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.