Top-down attention guidance shapes action encoding in the pSTS

Abstract

To test for the influence of top-down influences on action perceptual encoding, we evaluated the statistical structure of the multivariate activation pattern from the pSTS while observers attended to the different dimensions (the action kinematics, the goal, or the identity) of an avatar engaged in two different actions. Multivariate pattern decoding accuracy varied as a function of attention instruction in the right pSTS, but not in the other regions of the AON, with the highest classification when observers attended to the action kinematics. Furthermore, functional connectivity between the pSTS and inferior frontal cortex (IFC) was stronger when observers attended to the actions portrayed in the vignettes. Our findings are evidence that the attention goals of the viewer modulate sensory representations in the pSTS, which is proposals of the pSTS as an interstitial zone mediating top-down context and bottom-up perceptual cues during action observation.