Within the brain, the interplay between connectivity patterns of neurons and their spatiotemporal dynamics is believed to be intricately linked to the bases of behavior, such as the process of storing, consolidating, and retrieving memory traces. Memory is believed to be stored in the synaptic patterns of anatomical circuitry in the form of increased connectivity densities within subpopulations of neurons. At the same time, memory recall is thought to correspond to activation of discrete areas of the brain corresponding to those memories. Such regional subpopulations can selectively activate during memory recall or retrieval, signifying the process of accessing a single memory or concept. It has been shown previously that recovery of single memory activity patterns is mediated by global neuromodulation signifying transition into different cognitive states such as sleep or awake exploration. We examine how underlying topology can affect memory awake activation and sleep reactivation when such memories share increasing proportions of neurons. The results show that while single memory activation is diminished with increased overlap, pattern separation can be recovered by offsetting excitatory associations between two memories with targeted and heterogeneous inhibitory feedback. Such findings point to the importance of excitatory-to-inhibitory current balance at both the global and local levels in the context of memory retrieval and replay, and highlight the role of network topology in memory management processes.