Why X rather than Y? Explaining Neural Model' Predictions by Generating Intervention Counterfactual Samples

Open in new window