ALT a Schematized depiction of upright- and inverted-mapped mFus-faces pRF coverage when one views example faces centered on fixation. This predicts that viewing inverted faces yields suboptimal coverage of the location of the face features by pRFs in face-selective regions, which may contribute to the reported behavioral detriments in recognizing inverted faces (e.g., the face inversion effect, or FIE). b To evaluate the hypothesis schematized in a, we used a simulation procedure (see Methods) to determine the location in the visual field that would produce maximal pRF overlap with the internal features of 3.2° faces. To do so, we iteratively sampled a subset of mFus-faces voxels to generate bootstrapped coverage in response to inverted or upright faces; we then took the dot product of this bootstrapped coverage and a face centered at various retinotopic positions (central 3° × 3°, sampled at 0.1° steps). The retinotopic location that yielded the maximal overlap (dot product)
