Figure 4. A framework for content-specific beta activity. A, Content-specific beta-synchronization as endogenously driven transition from latent to active cortical representation. Left, Active cortical representations (e.g., of currently perceived, task-relevant information) are characterized by spiking activity (symbolized in red) in content-specific neuronal ensembles. Second from left, In the absence of perceptual input and/or attentional prioritization, information can be retained in latent memory representations, without spiking activity in the content-specific ensemble, e.g., in patterns of synaptic weights. Second from right, Endogenously driven (re)activation of a content-coding ensemble is characterized by a brief period of beta-synchronization, involving both local and long-range (top-down) interactions (see text for details). Right, (Re)activated content representations may again be characterized by spiking ensemble activity, similar (but not necessarily exactly identical) to representations of just perceived information (compare with left). B, Local beta activity appears content specific when population-level recordings register the synchronization of individual subpopulations (symbolized in blue and red) with differential sensitivity (e.g., red > blue, by spatial proximity to recording site). C, Transient network-level beta coherence in monkey PFC during application of different task rules (Fig. 2F, dashed rectangle). Adapted with permission from Buschman et al. (2012). Reprinted with permission from Elsevier.