Cortical Plasticity as a Mechanism for Storing Bayesian Priors in Sensory Perception
(a and c) Slopes of the input-output function (see the pink line in Figure 3D), showing the degree of prior bias. (b and d) Standard deviation of the decoded frequencies, which measures the decoding variability. In a and b, sensory-evoked activity level was fixed and the neuronal population size (color-coded) and baseline activity level were systematically varied. Baseline activity level refers to the multiplicative factor. For example, baseline activity level of 2 indicates doubling of activity. In c and d, the ratio of baseline activity to maximum evoked response magnitude was set at 1, and activity was systematically varied together. Error bars represent SEM, and are mostly masked by the data symbols.