Chong Zhao¹, Edward K. Vogel¹; ¹University of Chicago

Remembering a list of 128 real-world objects is much more challenging than memorizing a list with just a few objects. Previous behavioral studies had suggested that visual long term recognition tasks produce response times that scale logarithmically with list length. Here, we investigated the neural mechanisms underlying this phenomenon. Sixty participants studied lists of real-world objects with lengths of 1, 2, 4, 16, 32, and 128, followed by a recognition memory test while electroencephalograms (EEG) were recorded. Event-related potential (ERP) analyses revealed that the amplitudes of the parietal and frontal old-new effects significantly decreased as list length increased, particularly for lists of 32 and 128 items. Additionally, the latencies of these effects shifted systematically: shorter lists elicited earlier latencies, while longer lists exhibited progressively delayed responses. Critically, the latency increases were well-modeled by a logarithmic function of list length. These findings demonstrate that memory load modulates both the strength and timing of recognition-related neural activity, aligning with logarithmic memory search processes reported in hybrid search paradigms.

Acknowledgements: This research was supported by the funding from National Institute of Mental Health (grant ROIMH087214); Office of Naval Research (grant N00014-12-1-0972).