“…The use of metrics for the analysis of systems stems from [GJS90, DGJP04, KN96] where, in a process algebraic setting, it is argued that metrics are indeed more informative than behavioural equivalences when quantitative information on the behaviour is taken into account. The Wasserstein lifting has then found several successful applications: from the definition of behavioural metrics (e.g., [vB05,CLT20b,GLT16,GT18]), to privacy [CGPX14, CCP18, CCP20] and machine learning (e.g., [ACB17, GAA + 17, TBGS18]). Usually, one can use behavioural metrics to quantify how well an implementation (I) meets its specification (S).…”