Active Inference: A Process Theory
- PMID: 27870614
- DOI: 10.1162/NECO_a_00912
Active Inference: A Process Theory
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.
Similar articles
-
Neural Dynamics under Active Inference: Plausibility and Efficiency of Information Processing.Entropy (Basel). 2021 Apr 12;23(4):454. doi: 10.3390/e23040454. Entropy (Basel). 2021. PMID: 33921298 Free PMC article.
-
Morphogenesis as Bayesian inference: A variational approach to pattern formation and control in complex biological systems.Phys Life Rev. 2020 Jul;33:88-108. doi: 10.1016/j.plrev.2019.06.001. Epub 2019 Jun 12. Phys Life Rev. 2020. PMID: 31320316 Review.
-
Active inference leads to Bayesian neurophysiology.Neurosci Res. 2022 Feb;175:38-45. doi: 10.1016/j.neures.2021.12.003. Epub 2021 Dec 27. Neurosci Res. 2022. PMID: 34968557 Review.
-
Active inference and learning.Neurosci Biobehav Rev. 2016 Sep;68:862-879. doi: 10.1016/j.neubiorev.2016.06.022. Epub 2016 Jun 29. Neurosci Biobehav Rev. 2016. PMID: 27375276 Free PMC article. Review.
-
Universal Darwinism As a Process of Bayesian Inference.Front Syst Neurosci. 2016 Jun 7;10:49. doi: 10.3389/fnsys.2016.00049. eCollection 2016. Front Syst Neurosci. 2016. PMID: 27375438 Free PMC article.
Cited by
-
Repeating patterns: Predictive processing suggests an aesthetic learning role of the basal ganglia in repetitive stereotyped behaviors.Front Psychol. 2022 Sep 8;13:930293. doi: 10.3389/fpsyg.2022.930293. eCollection 2022. Front Psychol. 2022. PMID: 36160532 Free PMC article.
-
Linking actions and memories: Probing the interplay of action-effect congruency, agency experience, and recognition memory.Mem Cognit. 2024 Oct 9. doi: 10.3758/s13421-024-01644-2. Online ahead of print. Mem Cognit. 2024. PMID: 39382829
-
Joint modeling of choices and reaction times based on Bayesian contextual behavioral control.PLoS Comput Biol. 2024 Jul 5;20(7):e1012228. doi: 10.1371/journal.pcbi.1012228. eCollection 2024 Jul. PLoS Comput Biol. 2024. PMID: 38968304 Free PMC article.
-
Locus Coeruleus tracking of prediction errors optimises cognitive flexibility: An Active Inference model.PLoS Comput Biol. 2019 Jan 4;15(1):e1006267. doi: 10.1371/journal.pcbi.1006267. eCollection 2019 Jan. PLoS Comput Biol. 2019. PMID: 30608922 Free PMC article.
-
Model-based spatial navigation in the hippocampus-ventral striatum circuit: A computational analysis.PLoS Comput Biol. 2018 Sep 17;14(9):e1006316. doi: 10.1371/journal.pcbi.1006316. eCollection 2018 Sep. PLoS Comput Biol. 2018. PMID: 30222746 Free PMC article.
Publication types
LinkOut - more resources
Full Text Sources
Other Literature Sources
