1 paper · 1 filter
Yuliia Kaidashova, Bettina Finzel, Ute Schmid
Understanding why a classification model prefers one class over another for an input instance is the challenge of contrastive explanation. This work implements concept-based contra…