Fetching the paper…

Asymmetric Shapley values: incorporating causal knowledge into model-agnostic explainability · Around