Fig. 4: The three-level HGF binary perceptual model. | Schizophrenia

Fig. 4: The three-level HGF binary perceptual model.

From: Atypical prediction error learning is associated with prodromal symptoms in individuals at clinical high risk for psychosis

Fig. 4

An agent infers on multiple hidden states at time point k: stimulus probabilities (x1(k)), stimulus transition contingencies (x2(k)) and the phasic environmental volatility (x3(k)). Given new sensory input u(k), an agent updates their beliefs at each hierarchical level using simple one-step update equations of the sufficient statistics, the mean μ and the variance σ.

Back to article page