Due to transmission delays, the perceptual information our brain can access quickly becomes outdated as events unfold in real-time. We suggest our perceptual system learns internal representations that encode sequences (or timelines) rather than single points to compensate for transmission de- lays. Specifically, we investigate the dynamic predictive coding (DPC) model in which high-level states predict the transition dynamics of lower-level states and represent lower-level state sequences. We show that a two-level DPC network trained to predict videos captures several aspects of the well-known flash-lag illusion and exhibits both predictive and postdictive effects resembling those observed in human visual motion processing. Our results support the view that visual perception relies on temporally abstracted representations that encode sequences (or timelines) rather than single time points.

Methodological advances have made it possible to generatefMRI predictions for cognitive architectures, such as ACT-R, thus expanding the range of model predictions and mak-ing it possible to distinguish between alternative models thatproduce otherwise identical behavioral patterns. However, fortasks associated with relatively brief response times, fMRI pre-dictions are often not sufficient to compare alternative models.In this paper, we outline a method based on effective connec-tivity, which significantly augments the amount of informationthat can be extracted from fMRI data to distinguish betweenmodels. We show the application of this method in the caseof two competing ACT-R models of the Stroop task. Althoughthe models make, predictably, identical behavioral and BOLDtime-course predictions, patterns of functional connectivity fa-vor one model over the other. Finally, we show that the samedata suggests directions in which both models should be re-vised.