Martin Paré

The paper shows that models of visual working memory developed for single-report tasks fail to generalize to whole-report tasks due to assumptions about stimulus independence and encoding.

This paper develops a fronto-parietal cortical model to explain working memory overload by showing high-performing subjects avoid overload through competitive encoding and inter-areal projections.

This paper presents a neural network model of working memory that explains how memory capacity and coding fidelity decrease with increasing load through feedback inhibition.

The paper uses a cortical network model to show that VIP-mediated disinhibition controls flexible behavior by modulating working memory and decision-making.