Abstract:Highlights • Bayesian brain theory (BBT) is a conceptual framework in neuroscience that mathematically formalizes the dynamic of information processing through belief encoding and perceptual inference. • Contrary to the concept of belief as an explicit propositional attitude in psychology, BBT considers beliefs as probability densities about the latent causes of sensory data encoded in neural networks. • These probabilist beliefs are organized into cortical hierarchies to represent a multitude of spatiotemporal scales and causal relationships within a single probability space; • These probabilistic beliefs are constantly updated within synaptic plasticity, which is dependent on the timing of neuronal spikes (timing between pre- and postsynaptic activity) and glutamatergic excitation-inhibition balance. • Precision (inverse of the variance of a distribution) reflect the signal-to-noise ratio of sensory signals and is encoded by monoaminergic neuromodulators (serotonin, noradrenaline, dopamine). • Neural computation of perceptual inference requires computational resources that exceed the capacities of the brain, and it was proposed that neural networks compute variational Bayesian inference for belief updating. • The transition from Bayesian belief (implicit, subpersonal, probability density) to psychological belief (explicit, person-level, propositional attitude) is still unclear and needs to be clarified. Bayesian brain theory, a computational framework derived from the principles of Predictive Processing (PP), proposes a mechanistic formulation of belief generation and updating. This theory assumes that the brain encodes a generative model of its environment, made up of probabilistic beliefs organized in networks, from which it generates predictions about future sensory inputs. The difference between predictions and sensory signals produces prediction errors, which are used to update belief networks. In this article, we introduce the fundamental principles of the computational neuroscience of belief and show how this dynamic of prediction and updating offers a comprehensive account of the phenomenology of belief in psychiatry.

Computational Approaches to Access Probabilistic Population Codes for Higher Cognition an Decision-Making

A Radically New Theory of how the Brain Represents and Computes with Probabilities

Computational Neuropsychology and Bayesian Inference

Probabilistic vs. non-probabilistic approaches to the neurobiology of perceptual decision-making

Statistical Cognition Theory and Its Applications:An Overview

How does the brain compute with probabilities?

Predictive Coding: a Theoretical and Experimental Review

Neural Correlates of Optimal Multisensory Decision Making under Time-Varying Reliabilities with an Invariant Linear Probabilistic Population Code

A neural code supporting prospective probabilistic reasoning for instrumental information demand in humans

Learning probability distributions of sensory inputs with Monte Carlo Predictive Coding

A Bayesian brain model of adaptive behavior: An application to the Wisconsin Card Sorting Task

Toward a principled Bayesian workflow in cognitive science.

Bayesian brain theory: Computational neuroscience of belief

An Introduction to Predictive Processing Models of Perception and Decision-Making

Bayes in the Age of Intelligent Machines

A calculus of probability or belief? Neural underpinnings of social decision-making in a card game

Toward Biochemical Probabilistic Computation

Neuroscientific User Models: The Source of Uncertain User Feedback and Potentials for Improving Web Personalisation

Spike-based population coding and working memory

Neuroscientific User Models: The Source of Uncertain User Feedback and Potentials for Improving Recommendation and Personalisation

Neural Implementation of Bayesian Inference in Population Codes