Michael Epstein¹ (), Sophie Kovacevich¹, Rachel Denison¹; ¹Boston University

Background: Temporal interactions between successive stimuli occur across time ranges of hundreds of milliseconds. We previously found decreased perceptual sensitivity to targets that were preceded or followed by high-contrast vs. low-contrast non-targets, consistent with a principle of normalization across time. However, the mechanisms underlying these forward and backward temporal interactions are unknown. Goal: Here we used an orientation reproduction task to test whether perceptual suppression by temporal competitors is due to reduced orientation precision or a total loss of orientation information. Methods: Participants viewed a sequence of two Gabors (T1 and T2), each randomly oriented from 0-180 degrees. At the end of each trial, a response cue indicated the target, instructing participants to reproduce the orientation of either T1 or T2 using an adjustable probe. Stimuli were presented with 50 ms durations and a 250 ms stimulus onset asynchrony, with no temporal uncertainty, in the lower right screen quadrant. Contrasts were independently manipulated so that each target was either high (64%) or low (16%) contrast. Participants also reported if they missed (i.e., had no awareness of) one of the stimuli; missed trials were excluded from further analysis. Mixed models combining Gaussian and uniform error distributions were fit to the orientation estimation data using MemToolbox to determine how much of the perceptual suppression was explained by reduced precision vs. increased guessing. Results: Orientation estimates for both targets had greater error when paired with higher vs. lower contrast non-targets. Critically, the mixed modeling showed that these impairments arose by different mechanisms for each target. Higher non-target contrast predominantly reduced precision for T1 targets but predominantly increased guess rate for T2 targets. Conclusions: The findings suggest two distinct mechanisms for competition across time between successive stimuli, with implications for phenomena such as temporal crowding and models of dynamic perception.

Acknowledgements: Funding for this project was provided by a National Institute of Health National Eye Institute fellowship to M.E. (1F32EY033625), Boston University Undergraduate Research Opportunity Program funding to S.K. and Boston University startup funding to R.D.