📝 SummarXiv

Abstract

Bell inequalities rely on an assumption that the probabilities of adopting configurations of hidden variables describing a system prior to measurement are independent of the choice of measured physical property, also known as measurement independence. Weakening this assumption could alter the inequalities to accommodate experimental data whilst maintaining local interactions. A natural avenue for achieving this would be to model measurement as a dynamical process involving an interaction between the system and its environment (the measurement apparatus), that drives the hidden variables towards attractors representing measurement outcomes of the observable. Implementing such hidden variable dynamics, we can infer from observed correlations the hidden variable probability distributions before measurement, which differ according to which measurement settings were chosen. We explore various models of the dynamics of the hidden variables under measurement, revealing features that can create measurement dependence and others that can not.