Allen Institute Openscope - Loop project

In the natural world, sensory experience unfolds as a continuous and dynamic process where past sensory input predicts the present. The brain utilizes this sequential flow of recent sensory information to make predictions, infer causal relationships between events, and to give rise to our perceptions. Traditionally, such higher cognitive functions were attributed to the computations undertaken in associative brain regions, while sensory areas (such as V1) were thought to serve as passive feature detectors. This longstanding view is now being challenged by compelling evidence showing the capacity of these areas for plasticity and complex computations such as forming spatial expectations, temporal associations, reward prediction, and categorial learning. Importantly, it has been established that sensory processing is not solely determined by the system's current sensory inputs but also strongly influenced by preceding sensory inputs, a phenomenon known as hysteresis.This phenomenon has been characterized in detail in the visual system, with prominent examples that include adaptation, repetition suppression, prediction suppression and facilitatory history effects. However, previous work predominantly focused on the neuronal dynamics over short time scales, ranging from milliseconds to several minutes, in response to a limited set of synthetic stimuli. Moreover, much of our current knowledge relies on recordings from the primary visual cortex and on interpretations that emphasize modulation of single cells’ activity rather than analysis of population-level dynamics. Consequently, the extent to which prior experience over longer timescales shapes neuronal representations, and specifically representations of more ethologically relevant natural stimuli, remains unclear. It also remains unclear whether and to what extent the effects of prior experience on such neuronal representations differ along the visual system's hierarchical structure.