Maximally $\psi-$epistemic models cannot explain gambling with two qubits

Abstract

We investigate the minimal proof for ruling out maximally $\psi-$epistemic interpretations of quantum theory, in which the indistinguishable nature of two quantum states is fully explained by the epistemic overlap of their corresponding distributions over ontic states. To this end, we extend the standard notion of epistemic overlap by considering a penalized distinguishability game involving two states and three possible answers, named as Quantum Gambling. In this context, using only two pure states and their equal mixture, we present an experimentally robust no-go theorem for maximally $\psi-$epistemic models, showing that qubit states achieve the maximal difference between quantum and epistemic overlaps.