Information-theoretic constraints on correlations with indefinite causal order

TL;DR

This paper explores how quantum correlations with indefinite causal order can be constrained using information-theoretic principles, proposing new measures beyond mutual information for reconstructing causal structures.

Contribution

It introduces three principles within a generalized probabilistic framework that limit bipartite quantum correlations, advancing the understanding of causal structures in quantum theory.

Findings

Identifies principles constraining bipartite correlations to quantum bounds

Highlights the importance of alternative dependence measures beyond mutual information

Provides a framework for reconstructing causal structures in quantum theory

Abstract

Reconstructions of quantum theory usually implicitly assume that experimental events are ordered within a global causal structure. The process matrix framework accommodates quantum correlations that violate an inequality verified by all causally ordered correlations. Using a generalized probabilistic framework, we propose three principles constraining bipartite correlations to the quantum bound. Our approach highlights the role of a measure of dependence other than mutual information for an information-theoretic reconstruction of causal structures in quantum theory.