ForneyLab: A Toolbox for Biologically Plausible Free Energy Minimization in Dynamic Neural Models

Abstract

The free energy principle (FEP) claims that self-organization in biological agents is driven by variational free energy (FE) minimization in a generative probabilistic model of the agent’s environment [6]. Research progress in this field relies substantially on the capability to simulate biologically plausible FE minimization processes in postulated generative neural models. To this end, we have developed ForneyLab as a free and open source toolbox for FE minimization by variational message passing in freely definable probabilistic dynamic models. ForneyLab is released as a package for the open source scientific programming language Julia, which combines a MATLAB-like syntax and native speed close to that of (compiled) C code. ForneyLab is based on the formalism of Forney-style factor graphs (FGGs), which facilitates a very modular (plug-in) design approach and automated derivation of biologically plausible message passing-based inference algorithms.

Thijs van de Laar

Assistant professor

I am an assisant professor at BIASlab, where I work on artificial agents that learn to control themselves in uncertain environments. I take inspiration from physics and neuroscience, and develop theory and (software) tools that allow for efficient, real-time interaction.

Marco Cox

Former PhD student

Former researcher at BIASlab.

İsmail Şenöz

Chief Scientist
LazyDynamics

Ismail Senoz is a co-founder & chief scientist of Lazy Dynamics

Ivan Bocharov

Former PhD student

Former researcher at BIASlab.

Bert de Vries

Professor

I am a professor at TU Eindhoven and team leader of BIASlab.