Figure 5
From: Cyclic quantum annealing: searching for deep low-energy states in 5000-qubit spin glass
Distribution of clusters. Clusters are identified for final states from cycles with different annealing parameters \(b_x\). The parameter \(b_z\) is fixed as 0.03. (a) Power law distribution of cluster sizes. Data collapse for clusters of final states from cycles with different \(b_x\). The red line is fitted with power-law equation: \(\log (y) = -2.12 \log (x) -0.03\) with r-value − 0.95. The exponent 2.12 shows the variance is not finite, implying the possibility of rare super clusters. (b) Energy contribution \(\Delta E\) vs. Cluster Sizes. For each cluster, we calculated its energy contribution \(\Delta E\) to the energy difference with respect to the reference state. Clusters from cycles with different annealing parameters \(b_x\) are shown here in different color. The legend is the same as that in (a). Data of negative \(\Delta\)E are shown here. A dashed line shows the relation between energy decrease and smallest possible cluster size to achieve that. To find a large energy decrease, one must find a large cluster.
