We derive and compare device-independent witnesses of genuine multipartite entanglement (GME) and genuine multipartite nonlocality (GMNL) for GHZ states and trees of Bell pairs. A binary-outcome family using four global measurement contexts has the tight biseparable bound independent of the number of parties. We distinguish these GME bounds from Svetlichny-hybrid bounds, which permit arbitrary communication within either side of a bipartition. For GHZ states, we also give a
2N-context GMNL witness and compare it with MABK and Svetlichny tests. For Bell-pair trees, sums of CHSH and chained inequalities certify GME and GMNL, respectively, with explicitly specified input protocols and multi-bit local outputs. Separately, we prove the sharp four-context network bound
Iλ≤2 for every
N≥3 and
λ≥2, and bound the MABK scores of quantum networks assembled from bipartite sources. We successfully benchmark MABK, four-context, and Svetlichny witnesses on IBM Quantum GHZ preparations up to
10 parties, under the local-measurement and sampling assumptions.
We consider the problem of determining whether genuine multipartite entanglement was produced in an experiment, without relying on a characterization of the systems observed or of the measurements performed. We present an n-partite inequality that is satisfied by all correlations produced by measurements on biseparable quantum states, but which can be violated by n-partite entangled states, such as Greenberger-Horne-Zeilinger states. In contrast to traditional entanglement witnesses, the violation of this inequality implies that the state is not biseparable independently of the Hilbert space dimension and of the measured operators. Violation of this inequality does not imply, however, genuine multipartite nonlocality. We show more generically how the problem of identifying genuine tripartite entanglement in a device-independent way can be addressed through semidefinite programming.
In a recent paper, Bancal et al. [Phys. Rev. Lett. 106, 250404 (2011)] put forward the concept of device-independent witnesses of genuine multipartite entanglement. These witnesses are capable of verifying genuine multipartite entanglement produced in a laboratory without resorting to any knowledge of the dimension of the state space or of the specific form of the measurement operators. As a by-product they found a multiparty three-setting Bell inequality which makes it possible to detect genuine
n-partite entanglement in a noisy
n-qubit Greenberger-Horne-Zeilinger (GHZ) state for visibilities as low as
2/3 in a device-independent way. In this paper, we generalize this inequality to an arbitrary number of settings, demonstrating a threshold visibility of $2/\ensuremath{\pi}~0.6366$ for number of settings going to infinity. We also present a pseudotelepathy Bell inequality achieving the same threshold value. We argue that our device-independent witnesses are optimal in the sense that for
n odd the above value cannot be beaten with
n-party-correlation Bell inequalities.