Fig. 4: Assessing the task-specificity of spatial predictions.
From: The spatial layout of antagonistic brain regions is explicable based on geometric principles
Top: The approach consists of creating a mask from the intersection of two task-specific masks, predicting activation for one task across the rest of the cortex, and then assessing whether the predicted output is more similar to the (expected) true task than to the alternative task. Bottom: Pairwise similarity in true versus alternative task performance, based on: A All out-of-mask vertices; B Out-of-mask vertices further than 0.1 radians from any vertex in the predictor mask; C Out-of-mask vertices with empirical activity z > 0.
