Figure 5
(a) experimental \(\overline{\text {MI}}(k)\) for two different electrode couples to which \(\Delta V=1\) V is applied (1-A: azure and 1-C: orange). Data are collected over 275 distinct configurations. (b) \(\overline{\text {IN}}(k)\) for the same experimental data. Data are collected over 275 distinct configurations. In both cases, errorbars are given by the evaluation of these observables over all the different ways of choosing a given k. \(\overline{\text {MI}}\) is only measured up to \(k=5\), since \(\text {MI}(X_j^k,X-X_j^k)=\text {MI}(X-X_j^k,X_j^k)\), while \(\overline{\text {IN}}\) is measured for any \(k \in [0,9]\). Grey dashed lines represent the \(\overline{\text {IN}}\) growth in a fully correlated system.
