​​​​​Talk from  Laetitia Grabot, CNRS & ENS - Université PSL

Short abstract:​​

Perception depends on internal models of the world to interpret sensory information. These internal models - or prior knowledge - generate perceptual biases, often leading to distorted perceptions of the environment, thereby shaping each of our subjective realities. Two types of priors can be distinguished: contextual priors, driven by top-down influences from the environment, and structural priors, which likely arise from individual differences in brain anatomy. While contextual priors have been extensively studied, structural priors remain largely unexplored, with the influence of anatomical constraints on perception often overlooked. In my research, I focus on time perception—such as the perception of temporal order or duration—as a window into perceptual inference. Time perception is uniquely suited for this, being strongly influenced by both contextual and structural priors. My work has demonstrated the existence of a structural prior in time perception. Building on existing literature that links alpha oscillations (8–12 Hz) with contextual priors,
I have shown that these brain rhythms also correlate with structural priors. However, the precise mechanisms by which alpha oscillations relate to priors remain unclear. One emerging hypothesis is that alpha oscillations may travel across the cortex as traveling waves. Testing this hypothesis is challenging with non-invasive techniques like MEG and EEG due to their limited spatial resolution. To address this, I developed a model-based neuroimaging approach that enables the investigation of previously untestable hypotheses involving traveling waves. This approach aims to overcome current methodological limitations and will allow to explore the spatiotemporal brain dynamics underlying perceptual inference​

​