Quantum Non-classicality from Causal Data Fusion

Abstract

Bell's theorem, a cornerstone of quantum theory, shows that quantum correlations are incompatible with a classical theory of cause and effect. Through the lens of causal inference, it can be understood as a particular case of causal compatibility, which delves into the alignment of observational data with a given causal structure. Here, we explore the problem of causal data fusion that aims to piece together data tables collected under heterogeneous conditions. We investigate the quantum non-classicality that emerges when integrating both passive observations and interventions within an experimental setup. Referred to as "non-classicality from data fusion," this phenomenon is identified and scrutinized across all latent exogenous causal structures involving three observed variables. Notably, we demonstrate the existence of quantum non-classicality resulting from data fusion, even in scenarios where achieving standard Bell non-classicality is impossible. Furthermore, we showcase the potential for attaining non-classicality across multiple interventions using quantum resources. This work extends a more compact parallel letter on the same subject and provides all the required technical proofs.