Segregation-to-Integration Transformation Model of Memory Evolution

Abstract

AbstractMemories are thought to use coding schemes that dynamically adjust their representational structure to maximize both persistence and utility. However, the nature of these coding scheme adjustments and their impact on the temporal evolution of memory after initial encoding is unclear. Here we introduce the Segregation-to-Integration Transformation (SIT) Model, a neural network formalization that offers a unified account of how the representational structure of a memory is transformed over time. SIT Model asserts that memories initially adopt a highly modular or segregated network structure, functioning as an optimal storage buffer by striking a balance between protection from disruptions and accommodating substantial information. Over time, a repeated combination of neural network reactivations, spreading, and synaptic plasticity transforms the initial modular memory structure into an integrated memory form, facilitating intercommunity spreading and fostering generalization. In addition, SIT Model reveals the existence of an optimal window during this transformation where memories are most susceptible to malleability, suggesting a non-linear or inverted U-shaped function in memory evolution. The results of our model integrate a wide range of experimental phenomena along with accounts of memory consolidation and reconsolidation, offering a unique perspective on memory evolution by leveraging simple architectural neural network property rules.