Global explanations in captum?

are there any global explanations algorithms embedded in captum?
is there any plan to add some anytime soon?
which ones?

Hi @ccarmel, We are currently working onTCAV - Testing with Concept Activation Vectors.
Besides that we can always aggregate sample level attributions to get a global view, but this of course it tied to the data.
Do you have any global explanation algorithms for Neural Networks in mind ?