exactly what i have been arguing, for two decades (eg The Algebraic Mind) and in my 2018 critical appraisal of deep learning cc @stanfordnlp https://twitter.com/weihua916/status/1025137101040508931 …
This Tweet is unavailable.
-
Learned convolutional filters are certainly a representational advance. Word embeddings certainly capture some important regularities. What else would you consider to be interesting learned representations?
-
-
Has the learning of compositional semantics been demonstrated?
-
By children, every day.
End of conversation
New conversation -
-
-
Infinite use of finite means
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
The big thing is the ability to transform
#representations into other ones depending on the problem to solve. Mathematicians are doing this all the time. Logicians are doing this when switching between syntax and semantics.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.