The investigators recruited 14 adult patients undergoing evaluation at NYU Comprehensive Epilepsy Center prior to surgery, 24 healthy volunteers participating in-person, and 60 healthy volunteers participating online for experiments to examine the influence of long-term prior knowledge ('priors') on perception.

Bistable perception task: 14 epilepsy patients who were undergoing surgical evaluation with iEEG monitoring performed the bistable perception task. During this task, they were shown Necker cube and Rubin face-vase illusions. Participants were asked to passively view the images and report each time their percept changed by pressing one of two response buttons or ‘unsure’. iEEG was recorded from implanted electrodes that were placed according to clinical needs; participant #14 had an additional high-density grid implanted over the occipital lobe for research under a NIH BRAIN-funded project. 24 healthy volunteers also performed this task across three separate sessions spaced at least one week apart.

Online behavioral experiment: 60 healthy volunteers participated in this perception task. During each trial, participants were shown an instruction screen which informed them which image was about to be presented (i.e., Rubin’s face-vase or Necker cube), and the response mapping. After a two-second fixation period, an ambiguous image was presented on the screen for 60 seconds, while participants fixated on the cross in the center of the image, then pressed a key to indicate their current perception of the image. For the Rubin face-vase, the possible responses were face, vase, or "unsure". For Necker cubes, the possible responses were green, blue, or "unsure" to indicate the color of the face perceived as closest. The images were shown in three different sizes and four different locations for two trials.