Research questionHow can example-based explanations clarify how multiple differing features change an AI decision?A reference example can justify a prediction intuitively, but large differences across many attributes make it difficult to tell which changes drive the decision and by how much.