Quantum advantage in zero-error function computation with side information

We consider the problem of zero-error function computation with side information. Alice has a source $X$ and Bob has correlated source $Y$ and they can communicate via either classical or a quantum channel. Bob wants to calculate $f(X,Y)$ with zero error. We aim to characterize the minimum amount of information that Alice needs to send to Bob for this to happen with zero-error. In the classical setting, this quantity depends on the asymptotic growth of $\chi(G^{(m)})$, the chromatic number of an appropriately defined $m$-instance "confusion graph". In this work we present structural characterizations of $G^{(m)}$ and demonstrate two function computation scenarios that have the same single-instance confusion graph. However, in one case there a strict advantage in using quantum transmission as against classical transmission, whereas there is no such advantage in the other case.
Comment: We have realized an error in Claim 3