The surprising thing about deep learning isn't that our models can fit to basically any dataset, it is that they generalize --sometimes.
The key to generalization is abstraction. To understand DL, we need mathematical models of abstraction itself. An abstraction of abstraction
-
-
2deep4me
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I've been re-reading Hinton and
@OriolVinyalsML's paper on Model Distillation, and have gotten a bit of intuition about this topic recently.pic.twitter.com/evmqq2HuM6
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
that math is category theory. bt key to abstraction is learning in terms of generators. nothing of like yet. prog learning on way
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
cough cough category theory
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Abstraction is a transient phenom; important (IMO) not to reify abstracta.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Or is it the opposite: the key to abstraction is generalization
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Simple arithmetic is an abstraction of abstraction since numbers are an abstraction.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
how high are you?
-
I'm on a different kind of high. It's called research.
End of conversation
New conversation -
-
-
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.