Paul S. Scotti

This paper introduces an enhanced inverted encoding model (eIEM) for reconstructing perception, attention, and memory from neuroimaging data, showing improved reconstruction quality and brain-behavior correlations.

This study uses fMRI and an inverted encoding model to show that task-relevant features of attended objects are reliably reconstructed, while task-irrelevant features show weaker and less reliable reconstructions in visual and parietal cortex.