This is gorgeous. And it is chipping away at an important (& extremely difficult) problem, trying to understand how & why neural networks work.https://twitter.com/distillpub/status/1103339491589124099 …
Show this thread
I rather like the idea of neural nets as this interesting new phenomenon in the world, which may well be governed by underlying principles which are hidden, but which we can figure out. (Reminded of Herb Simon on the sciences of the artificial.)
-
-
From another angle, here's a remarkable recent paper that shows that, in the infinite width limit, the dynamics of neural nets learning under gradient descent can be solved *analytically*: https://arxiv.org/abs/1806.07572
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Been thinking about this article a lot, w.r.t. our new poisoned-stones-electrified-with-symbols friends:https://jods.mitpress.mit.edu/pub/lewis-arista-pechawis-kite …
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
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.