Fast Axiomatic Attribution for Neural Networks - Supplemental Material - Robin Hesse

In the proof of Proposition 3.2, we make use of the property that the derivative of a ( k 1) If the pooling function is linear, homogeneity implicitly holds. If the pooling function is selecting values based on their relative ordering, we consider two cases. For Expected Gradients to satisfy the same axioms that are satisfied by Integrated Gradients, convergence must have occurred, which can only be expected after multiple gradient evaluations. Gradient, Sensitivity (a) is also not satisfied by Expected Gradients in general. Sensitivity (b): As the gradient w.r .t. an irrelevant feature will always be zero, Sensitivity (b) Why is nonnegative homogeneity a desirable axiom for attribution methods?