ELECTROPHYSIOLOGICAL SIGNATURES OF THE CHANGE BIAS REVEAL HOW TIME PERCEPTION DEPENDS ON STIMULUS CONTENT AND SENSORY RESPONSE MAGNITUDE

When perceiving sensory input, human observers extract a sense of duration alongside other stimulus content. Stimulus timing is tightly interwoven with sensory processing. For example, stimuli that cause increased responses in sensory neurons are often found to be perceived as longer. Yet, where along sensory pathways perceived duration is computed remains unknown. Here, we present two EEG studies on how response magnitude for specific visual features relates to perceived time. We rely on the 'change bias', a paradigm where participants estimate an interval indicated by two briefly presented visual stimuli (start- and end marker). When markers are identical, neural repetition suppression attenuates sensory responses compared to trials where visual features change between markers. The change bias effect entails that intervals with a change are also perceived as longer. Previously, we have shown that the change bias is selective: certain feature changes (e.g., size, location) modulate time perception but others (e.g., orientation) do not. Across two studies we replicate this finding, and show that sensory responses to the end marker are indeed modulated by repetition suppression, correlating with the behavioral change bias. Furthermore, we show that offset responses can be classified as a repetition or change, but only in conditions with a behavioral effect. Decoding scores predicted effect size across participants. Finally, we could decode duration from offset responses and use decoding scores to predict behavioral precision. The results suggest that stimulus-specific repetition suppression can be leveraged to dissect which sensory pathways relate to time perception.