Schedule of Events | Search Abstracts | Symposia | Invited Symposia | Poster Sessions | Data Blitz Sessions

Poster F35

Perceptual and neural modulations by regularity of different information value

Poster Session F - Tuesday, April 16, 2024, 8:00 – 10:00 am EDT, Sheraton Hall ABC

Joey Zhou1 (, Sage Boettcher1, Kia Nobre1,2; 1University of Oxford, 2Yale University

The brain is sensitive to regularities in the environment and can use them to make perceptual predictions. Previous work on how these predictions modulate sensory neural processes reported mixed findings. While some studies showed that predicted/expected stimuli are associated with attenuated sensory response compared to the unexpected, others observed enhanced sensory response to the expected. Here, we aimed to reconcile these conflicting results by testing a novel hypothesis that regularities modulate perception differently depending on their information value. Participant’s task was to compare two consecutively presented targets (S1 and S2) in either their orientation or location, depending on the task cue presented at trial start. In half of the trials, S1 location is 100% predictable by a preceding cue. For orientation judgements, the cue-S1 regularity is of high information value, as it allowed participants to direct their attention to enhance S1 discrimination. For location judgements, however, the regularity is of low information value, because interpreting the cue alone is sufficient. In Experiment 1, we tested participants online and validated that valid cue-S1 regularity led to better accuracy in both tasks. In Experiment 2, we titrated S1’s contrast to match both tasks in their difficulty, and had participants performed the same task while recording their brain activity using MEG. By comparing the evoked activity of S1 when it was predicted vs. not predicted by the cue, and by examining whether such an effect differs between tasks, our results provide new understandings of predictive perception.

Topic Area: ATTENTION: Spatial


CNS Account Login


April 13–16  |  2024