Global Concept Explanations for Graphs by Contrastive Learning

Jonas Teufel, Pascal Friederich

Beyond improving trust and validating model fairness, xAI practices also have the potential to recover valuable scientific insights in application domains where little to no prior…