D. Rahnev

Human-like dissociations between confidence and accuracy in convolutional neural networks

This study shows that convolutional neural networks exhibit confidence-accuracy dissociations similar to humans when stimulus energy is manipulated, suggesting these effects arise from low-level representation changes rather than cognitive heuristics.

Distinguishing the roles of dorsolateral and anterior PFC in visual metacognition

This study uses TMS to show that DLPFC and aPFC have distinct roles in confidence generation during visual metacognition, with DLPFC lowering confidence and aPFC enhancing it in later task blocks.