Abstract
Experience with visual objects leads to later improvements in identification speed and accuracy (“repetition priming”), but generally leads to reductions in neural activity in single-cell recording studies in monkeys and fMRI studies in humans (“repetition suppression”). While the cell mechanisms that lead to these activity reductions are unclear, previous studies have implicated relatively local, automatic cortical mechanisms, and slice physiological recordings have identified several candidate short- and long-term plasticity mechanisms. I will show that these plasticity mechanisms when incorporated into a simplified neocortical circuit model are capable of re-producing changes in stimulus selectivity due to repetition as seen in single-cell recording studies in monkey area TE: “scaling” with relatively short-term repetitions and “sharpening” over longer periods of experience. However, these simulations when based on average firing rate fail to provide an account of behavioral priming. In contrast, simulations that retain the spiking property of neurons can potentially account for both repetition suppression and priming by allowing more synchronized and temporally coordinated activity at lower overall rates. I will review the current state of evidence in support of this proposal from monkey single-cell and LFP recordings and human MEG. I will also present new data from intracranial EEG recordings of human epilepsy patients showing that stimulus repetition at both short and long time scales leads to larger amplitude activity fluctuations at low frequencies (<15 Hz). These results indicate that greater neural synchronization accompanies lower overall activity levels following stimulus repetition, constituting a novel efficiency mechanism.