Testing the effectiveness of saliency-based explainability in NLP using randomized survey-based experiments

Open in new window