Project Summary The locations where our life experiences unfold are remarkably potent and enduring cues to memory. For instance, stepping back into a familiar café can reignite the distant memory of a casual coffee chat with a colleague. What then, are the properties of a spatial context and its neural representation that enable it to become a trigger to episodic memory? As we age our memories of the encoding context in which events occur deteriorate and can dramatically worsen in diseases like Alzheimer’s. Therefore, developing an understanding of how spatial contexts influence our memory during encoding and retrieval of naturalistic events is fundamental towards developing treatments to memory impairment in diseases like Alzheimer’s and dementia. Additionally, the enhanced accessibility, portability, and growing affordability of immersive virtual reality (VR) technology have made it an valuable asset for investigating how spatial contexts influence memory formation in more ecologically valid experiments. Thus, this proposal seeks to mechanistically understand how spatial contexts scaffold memories through an innovative research program that combines electroencephalography (EEG), VR, and functional resonance magnetic imaging (fMRI) in naturalistic tasks. A prevalent theory in Psychology posits that memories need to be both stable over time and distinct enough to prevent interference with other potentially similar memories. To what extent is this true for spatial context memories and how do these properties influence their role as the containers of subsequent memories that occur in them? In Aim 1, we will characterize the representational properties of a spatial context and how they transform across learning episodes. We hypothesize that across learning spatial context representations become increasingly more dissimilar to other contexts and that their stability (operationalized as increased pattern self-similarity across a delay) will converge to a stable state. In Aim 2, we will combine simultaneous VR and EEG in a real-time neurofeedback paradigm to directly distort the neural representations of spatial contexts to uncover their role as encoding contexts for subsequent memories. We predict that enhancing or reducing how distinct spatial contexts are from one another will improve or impair memory formation associated with events occurring in those spaces. In Aim 3, we will first identify what properties of image sequences make scenes more recallable over others, identify the neural correlates (in fMRI) that are involved in predicting their recallability, and then, develop a computational model that generates scenes along increasing or decreasing recallability. Taken together, the proposed project makes use of innovative, affordable, and accessible technology (i.e., VR-EEG) to study the fundamental nature of how spatial contexts influence memory for everyday events. In the long term, the findings and methodologies developed will support future ...