-
-
- Show replies
New conversation -
-
-
Looks like a ‘textbook example’ of NN learning spurious correlations that do not generalize to new data
-
On the dangers of spurious correlations: https://ff13.fastforwardlabs.com/#causality-and-invariance …
End of conversation
New conversation -
-
-
I noticed that art generators using clip also tend to produce a lot of logos because the encoders are rewarded for image to text in the dataset
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Agreed but I have a much stronger opinionhttps://twitter.com/A_K_Nain/status/1421861216004153344?s=19 …
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
The font of AI knowledge
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I thought "data leakage" is presence of Val data in train data only. What are different types of "data leakage"?
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.