Active Inference: A Process Theory

TLDR

The article presents a process theory grounded in active inference and belief propagation, noting that neuronal dynamics can be viewed as gradient descent on variational free energy, making it a Lyapunov function that aligns with Hamilton’s principle of least action. The study investigates whether neuronal responses can be modeled as gradient descent on variational free energy, based on the premise that all neuronal processing maximizes Bayesian model evidence. By applying a standard Markov decision process generative model, the authors derive neuronal dynamics that account for a wide array of established neuronal phenomena. The derived dynamics reproduce numerous phenomena—including repetition suppression, mismatch negativity, violation responses, place‑cell activity, phase precession, theta sequences, theta‑gamma coupling, evidence accumulation, race‑to‑bound dynamics, and dopamine response transfer—and provide a formal explanation for reward seeking, context learning, and epistemic foraging.

Abstract

This article describes a process theory based on active inference and belief propagation. Starting from the premise that all neuronal processing (and action selection) can be explained by maximizing Bayesian model evidence-or minimizing variational free energy-we ask whether neuronal responses can be described as a gradient descent on variational free energy. Using a standard (Markov decision process) generative model, we derive the neuronal dynamics implicit in this description and reproduce a remarkable range of well-characterized neuronal phenomena. These include repetition suppression, mismatch negativity, violation responses, place-cell activity, phase precession, theta sequences, theta-gamma coupling, evidence accumulation, race-to-bound dynamics, and transfer of dopamine responses. Furthermore, the (approximately Bayes' optimal) behavior prescribed by these dynamics has a degree of face validity, providing a formal explanation for reward seeking, context learning, and epistemic foraging. Technically, the fact that a gradient descent appears to be a valid description of neuronal activity means that variational free energy is a Lyapunov function for neuronal dynamics, which therefore conform to Hamilton's principle of least action.