Message passing-based inference in an autoregressive active inference agent
Abstract
We present the design of an autoregressive active inference agent in the form of message passing on a factor graph. Expected free energy is derived and distributed across a planning graph. The proposed agent is validated on a robot navigation task, demonstrating exploration and exploitation in a continuous-valued observation space with bounded continuous-valued actions. Compared to a classical optimal controller, the agent modulates action based on predictive uncertainty, arriving later but with a better model of the robot's dynamics.
Cite
@article{arxiv.2509.25482,
title = {Message passing-based inference in an autoregressive active inference agent},
author = {Wouter M. Kouw and Tim N. Nisslbeck and Wouter L. N. Nuijten},
journal= {arXiv preprint arXiv:2509.25482},
year = {2026}
}
Comments
14 pages, 4 figures, proceedings of the International Workshop on Active Inference 2025. Erratum v1: in Eq. (50), $p(y_t, \Theta, u_t \mid y_{*}, \mathcal{D}_k)$ should have been $p(y_t, \Theta \mid u_t, y_{*}, \mathcal{D}_k)$