2 ms·
> Apparently their approach can be used to scaffold any biased classifier in a manner that its predictions on the inputs remain biased but post hoc explanations
by AnthonyMouse 3y ago
> Apparently their approach can be used to scaffold any biased classifier in a manner that its predictions on the inputs remain biased but post hoc explanations come across as fair.
Passes the Turing test.