interestingly, attention didn't change the total size of the space, compared to perceptual judgements we measured without the interleaved attention task, consistent with normalization of the representations. we think this is a cool and novel way to look at the effects of attention - hope you agree!

we measured the local length throughout the orientation space - the distance spanned in windows of the representation - to assess changes induced by attention. lengths were expanded in a narrow window around targets, but compressed in a broad window around distractors.

our participants performed interleaved tasks, where they reported the perceived similarity between orientations under different attention conditions (attending 45° CW or CCW). we fit these responses with models based on MDS, finding that orientation representations were best accounted for in 4-D.