Show Thoughts

According to Knill and Pouget, being an optimal Bayesian observer only means to take into account the uncertainty of the available information (in the system—that's after lossy transformation from physical stimuli to neural representations).