A new tool for explainable AI

Explaining models trained in Julia, Python and R through counterfactuals

Turning a 9 (nine) into a 4 (four). Image by author.

Counterfactual explanations, which I introduced in one of my previous posts, offer a simple and intuitive way to explain black-box models without opening them. Still, as of today there exists only one open-source library that provides a unifying approach to generate and benchmark counterfactual explanations for models built and trained in Python (Pawelczyk et al. 2021). This is …

ai deep learning editors pick explainable ai julialang tool