Self-testing of Nonmaximal Genuine Entangled States using Tripartite Hardy Relations (2504.19491v1)

Abstract: We demonstrate that, in the tripartite scenario with all parties' local events being space-like separated, Hardy-type nonlocality constitutes a stronger manifestation of nonlocal correlations than those captured by Mermin-type inequalities, an important distinction that has hitherto remained unrecognised. To substantiate this assertion, we develop a general framework for the characterisation of tripartite correlations by extending the notion of Settings Independence and Outcome Independence beyond their bipartite formulation. This framework highlights the pivotal role of Hardy-type reasoning in the detection and certification of genuine multipartite nonlocality. Furthermore, we show that the tripartite Hardy-nonlocality enables the self-testing of a broad class of pure nonmaximally genuine entangled tripartite states. A key advantage of Hardy-based self-testing over methods based on tripartite Bell inequalities is its ability to certify quantum correlations even in the presence of nonmaximal violations. This, in turn, facilitates the device-independent certification of randomness from Hardy-type correlations. Unlike Bell functionals, which typically enable self-testing of only a single extremal point per inequality, Hardy relation self-tests a set of extremal quantum correlations for any nonzero Hardy probability. We find that the maximum certifiable randomness using Hardy-type correlations is $\log_2 7\approx 2.8073$-bits, highlighting both the practical and foundational significance of Hardy-based techniques for quantum randomness generation.