Mechanisms supporting superior source memory for familiar items: A multi-voxel pattern analysis study

► Classifier read-out of on-task processing predicted later memory for task. ► Previously known proverbs evoked greater on-task processing than novel or repeated ones. ► On-task processing was a better memory predictor for repeated proverbs than other types. ► These effects help explain memory bonuses of pre-study repetition and prior knowledge. ► Our results suggest a possible dissociation in the mechanisms underlying such memory bonuses.