Bayesian mechanics of perceptual inference and motor control in the brain

TL;DR

This paper develops a Bayesian mechanics framework based on the free energy principle, modeling active perception and motor control in the brain through Hamiltonian dynamics and variational Bayes.

Contribution

It introduces a physically principled formalism for the free energy principle by implementing free energy minimization via least action, extending to active inference.

Findings

Demonstrates recognition dynamics in a simple agent model

Provides a Hamiltonian formulation of neural active inference

Compares new approach with traditional state-space models

Abstract

The free energy principle (FEP) in the neurosciences stipulates that all viable agents induce and minimize informational free energy in the brain to fit their environmental niche. In this study, we continue our effort to make the FEP a more physically principled formalism by implementing free energy minimization based on the principle of least action. We build a Bayesian mechanics (BM) by casting the formulation reported in the earlier publication (Kim 2018) to considering active inference beyond passive perception. The BM is a neural implementation of variational Bayes under the FEP in continuous time. The resulting BM is provided as an effective Hamilton's equation of motion and subject to the control signal arising from the brain's prediction errors at the proprioceptive level. To demonstrate the utility of our approach, we adopt a simple agent-based model and present a concrete numerical illustration of the brain performing recognition dynamics by integrating BM in neural phase space. Furthermore, we recapitulate the major theoretical architectures in the FEP by comparing our approach with the common state-space formulations.