Schedule of Events | Symposia

Sketchpad Series

Complexity compression and the when of memory in virtual reality, an EEG study

Poster Session C - Sunday, March 30, 2025, 5:00 – 7:00 pm EDT, Back Bay Ballroom/Republic Ballroom

Matthew Logie¹ (matthewrlogie@gmail.com), Virginie van Wassenhove¹; ¹CEA/DRF/Inst. Koliot, Neurospin; INSERM, Cognitive Neuroimaging Unit; Université Paris Saclay, Gif/Yvette, 91191 France

Memory for the temporal distance between items depends on the structure of events. Events depend on contextual drift, which ‘resets’ when experiencing event boundaries. Boundaries are traditionally defined in terms of prediction error. An event is constructed within working memory along with predictions for what comes next. Prediction errors act as anchors for the compression and encoding of events into longer-lasting memory and clear the contents of working memory in preparation for the next event. The construction of an event can also be described in terms of an accumulation of information. Here we take an information theoretic approach and aim to explore the construction, compression and encoding of events in terms of an accumulation of information. We conducted an EEG study employing a novel virtual reality environment to provide segmented sequences of images with transitions between virtual locations providing boundaries to ‘reset’ the accumulation of information. From this perspective, we expect to discover a waxing and waning in measures of complexity and entropy based on event structure. Preliminary findings suggest that the level of entropy will influence the compression of information, which will drive temporal displacement errors in behavioural judgements of when an image appeared. We aim to establish links between the processing of information and the when of memory. Keywords: Virtual reality, EEG, Events, Context drift, Complexity

Topic Area: LONG-TERM MEMORY: Episodic