Eye movements show clear systematic patterns only for the active condition in some participants. (a) The top row shows the eye position over the course of a trial in the active session. (b) The bottom row shows the eye position over the course of a trial in the passive session. Colours represent trials belonging to one of eight orientations, and are those as shown in Fig. 1a. The solid lines represent averages over repeated presentations of the same orientation. The dashed line represents the inner annulus of the square-wave gratings, which had a radius of 1.5 visual degrees. Additionally, averages are computed for those participants with higher than 30% classification accuracy in the active session (N = 4) shown in the first column, those with lower than 30% classification accuracy but still significant decoding in the active session (N = 12) in the middle column, and those that were not decodable in the active session (N = 12) in the right column. Note that the passive session was never decodable, but for comparison the same groups are shown. This figure illustrates that for some participants there is a clear pattern in their eye movements, while for others this pattern is less clear. Additionally, these patterns remain completely absent in the passive sessions, even for those showing large patterns in the active session.