Karol P. Szymula

The paper investigates how neural responses differ during learning of modular and non-modular graph structures using fMRI data, finding distinct connectivity patterns associated with performance on different graph types.

The paper shows that the structure of a temporal sequence affects how well stimuli are represented in the brain, with modular graphs leading to better prediction and higher dimensionality in visual areas.

This study links cognitive map formation during sequence learning to neural activity in the temporal lobe using behavioral modeling and iEEG recordings.