Fig. 3

The outcome of the predictive coding process depends on weight (precision) attribution. On the left: when priors are judged to be more precise than incoming sensory information (likelihood), the posterior will be primarily based on priors (i.e., the posterior is shifted towards priors and there is no updating of existing beliefs); on the right: if the likelihood is judged to be more accurate than priors, the posterior will be mainly based on incoming sensory information (i.e., the posterior is shifted toward priors and there is belief update; prior blinding; image from [122])