Macroscopic limit of nonclassical correlations

TL;DR

This paper demonstrates that macroscopic measurements on large bipartite systems described by generalized probabilistic theories, including PR-boxes, can be explained by local hidden variables, challenging previous contradictory results.

Contribution

It proves that for large N, macroscopic correlations in such systems admit local hidden variable explanations, extending previous results to include fluctuations and more measurement settings.

Findings

Existence of joint probability distributions for large N

Macroscopic measurements reveal local hidden variables

Results challenge previous contradictory findings

Abstract

We consider macroscopic correlations in a bipartite system consisting of 2N particles described by a generalised probabilistic theory. In particular, we discuss a case of N PR-boxes shared between two parties. We characterise macroscopic measurements as collective measurements of the same property on all the boxes in the same region. Such measurements are assumed to reveal only the average value of the measured collective property. We show that for two measurements per observer and N greater or equal 2 there always exist a joint probability distribution explaining all observable data and therefore the system admits local hidden variables. Next, we generalise this result to include measurement of fluctuations and additional measurement settings. Finally, we discuss our result in the context of previous works in which contradictory results were presented.