, 2009 and Law and Gold, 2009), a more prevalent framework to study perception has been the “Bayesian brain hypothesis” that the brain constructs and updates a generative

model of its sensory inputs (Doya et al., 2011). One particular formulation of this hypothesis is predictive coding (Friston, 2005 and Rao and Ballard, 1999) that postulates that PEs are weighted by their precision and are computed at any level of hierarchically organized information processing cascades, as in sensory systems. This has been examined by several fMRI studies that contrasted predictable versus unpredictable visual stimuli, finding PE responses in visual areas specialized for the respective stimuli used (Harrison et al., 2007 and Summerfield and Koechlin, 2008) and precision-weighting under attention (Kok et al., 2012). Other studies have used an explicit model of trial-wise PEs, using visual (Egner et al., Volasertib mouse 2010) or audio-visual associative learning (den Ouden et al., 2010 and den Ouden et al., ABT-888 solubility dmso 2009) paradigms. Notably,

these studies did not have explicit readouts of subjects’ predictions and used relatively simple modeling approaches: they either described implicit learning processes (in the absence of behavioral responses) using a delta-rule RL model (den Ouden et al., 2009 and Egner et al., 2010), or dealt with indirect measures of prediction (e.g., reaction times) using an ideal Bayesian observer with a fixed learning trajectory across subjects (den Ouden et al., 2010). Our

present study goes beyond these previous attempts by (1) requiring explicit trial-by-trial many predictions, and (2) characterizing learning via a hierarchical Bayesian model that provides subject- and trial-specific estimates of precision-weighted PEs at different hierarchical levels of computation. Based on these advances, the present study shows much more widespread sensory PE responses than previously reported. Replicated in two separate groups, these responses were not only found in the visual cortex, but also in many supramodal areas in prefrontal, cingulate, parietal, and insular cortex (Figure 2). Whereas a distribution of reward (Vickery et al., 2011) and value signals (FitzGerald et al., 2012) across the whole brain have recently been demonstrated in humans, this has not yet been shown, to our knowledge, for PEs; in this case, precision-weighted PEs about the sensory outcome (visual stimuli). Perhaps the most interesting aspect of our findings on sensory outcome PEs, ε2, was the significant activation of the midbrain. In humans, strong empirical evidence exists for DA involvement in processing reward PEs (Montague et al., 2004 and Schultz et al., 1997) and novelty (Bunzeck and Düzel, 2006).