Confidence Enhances the Decodability of Predictions in Parietal Cortex During Spatial Decision-Making

Abstract

The ability to make predictions often involves some degree of uncertainty—a key determinant of confidence. Here, we aimed to assess whether predictions are decodable in partially-observable environments where uncertainty exists regarding the one’s state, and if this information is sensitive to confidence produced by state uncertainty. We employed functional magnetic resonance imaging based, partially-observable maze navigation task in which subjects predict upcoming scenes and report their confidence regarding these predictions. Using a multi-voxel pattern analysis, we successfully decoded both scene predictions and subjective confidence from activities in localized parietal and prefrontal regions. We also estimated the confidence in their belief about where they were in the maze. Importantly, prediction decoding accuracy varied according to subjective scene confidence and estimated state confidence in the superior parietal lobule and supramarginal gyrus, respectively. These results suggest that prediction in uncertain environments depends on the prefrontal-parietal network within which prediction and confidence interact.