Fig. 1: Understanding O-information in the context of the TSE complexity.
From: Multivariate information theory uncovers synergistic subsystems of the human cerebral cortex
The TSE complexity provides a system-wide summary statistic of how integrated the system is at every scale. The O-information can be understood as measuring how sensitive the global integration is to the removal of single elements (on average). a The panel shows a TSE curve for a low-synergy system: TC(X)/N > DTC(X)/N, so Ω(X) > 0. The erasure of any single element, on average, does not change the overall integration of the remaining (N − 1) elements much more than would be expected in the null case. b The panel shows the case where TC(X)/N < DTC(X)/N, so Ω(X) < 0 and the system is synergy dominated. Intuitively, this can be understood by recognizing that, on average, the removal of any of the N elements causes a large decrease in the integration among the remaining (N − 1) elements.
