Lara C. Krisst¹ (), Ailene Y. C. Chan¹, Noelle R. B. Stiles^1,2, Carmel A. Levitan³, Shinsuke Shimojo¹; ¹California Institute of Technology, ²Rutgers University, ³Occidental College

Previous research on multisensory integration reveals that participants often perceive an ‘illusory’ third visual stimulus when two visual flashes are presented at different locations, with three auditory tones (Stiles et al., 2018; Audio-visual Rabbit Illusion (AV Rabbit)). This crossmodal interaction was shown to be postdictive (wherein a later stimulus influences perception of an earlier stimulus). What is yet unknown is how this type of integration is represented in the brain. To this end, we use EEG in combination with multivariate pattern analysis to track the latency of the visual representations during the classic AV Rabbit paradigm. Participants are shown flashes moving along a horizontal/vertical trajectory and asked to report how many visual flashes they perceived. In the current task, participants are presented with two teardrop-shaped stimuli that rotate around a central origin, resembling the motion of clock hands. If participants perceive three teardrops (i.e., two real, one illusory in-between), they are instructed to report the orientation of the second (illusory) teardrop in a 360° space. We decoded the orientation information at each moment in time and compared trials in which participants reported seeing a third ‘illusory’ flash versus when they perceived only two flashes. Trials were physically identical (i.e., 3 beeps, 2 flashes), except for differences in the presence or absence of the ‘illusory’ flash. If the ‘illusory’ stimulus is postdictive in nature, we might expect this to influence the decoding accuracy of the visual stimuli. AV Rabbit behavioral effects were replicated with the teardrop-stimulus. Preliminary EEG findings show differences in decoding accuracy between the trials in which illusions were perceived and those in which perception was veridical, with earlier increases in decoding accuracy on illusory trials. Further investigation of the temporal dynamics of crossmodal illusory perception is required to fully elucidate mechanisms of postdiction in the brain.