How does the accuracy of sensory representations constrain the ability of the organism to perform perceptual discriminations? In this project, we will address this issue by characterizing the activity of populations of neurons in the rat auditory cortex while they perform a two-alternative forced-choice discrimination task for detecting sound location relative to the midline. We will first establish how location and spectral features of sound are represented by cortical populations under anesthesia, and subsequently during performance of the sound localization task. Due to the design of our discrimination task, analysis of the patterns of co-variability in the recorded sensory responses will allow us to elucidate whether the auditory cortex is primarily concerned with maximizing information about auditory stimuli, or whether it constructs representations which are useful for the animal in the context of its current behavioral goals. We will also investigate whether population activity and behavioral choices are correlated in a trial-by-trial basis. By understanding how the activity of multiple neurons has to be combined in order to best predict behavioral choices we expect to learn about the computations that the auditory cortex is performing in order to classify the location of sounds relative to the midline.
Fields of science
Call for proposal
See other projects for this call