A Message Passing Realization of Expected Free Energy Minimization

TL;DR

This paper introduces a message passing method for minimizing Expected Free Energy on factor graphs, transforming a complex search into a tractable inference problem, and demonstrates its effectiveness in uncertain environments.

Contribution

It reformulates EFE minimization as variational free energy minimization with epistemic priors, enabling efficient policy inference through standard variational techniques.

Findings

Agents outperform conventional KL-control agents in uncertain environments

EFE-minimizing agents avoid risky paths in stochastic gridworld

Agents conduct more systematic information-seeking in Minigrid tasks

Abstract

We present a message passing approach to Expected Free Energy (EFE) minimization on factor graphs, based on the theory introduced in arXiv:2504.14898. By reformulating EFE minimization as Variational Free Energy minimization with epistemic priors, we transform a combinatorial search problem into a tractable inference problem solvable through standard variational techniques. Applying our message passing method to factorized state-space models enables efficient policy inference. We evaluate our method on environments with epistemic uncertainty: a stochastic gridworld and a partially observable Minigrid task. Agents using our approach consistently outperform conventional KL-control agents on these tasks, showing more robust planning and efficient exploration under uncertainty. In the stochastic gridworld environment, EFE-minimizing agents avoid risky paths, while in the partially observable minigrid setting, they conduct more systematic information-seeking. This approach bridges active inference theory with practical implementations, providing empirical evidence for the efficiency of epistemic priors in artificial agents.