Belief-Based Maximum Occupancy Principle and Active Inference

Abstract

Intrinsic motivation plays a central role in adaptive and goal-directed behavior by conferring agents reward-independent objectives and biases useful to act in noisy and uncertain environments. Active Inference addresses the problem of acting in a partially observable environment through a principled framework for belief updating and action selection. A key component of Active Inference is the specification of prior preferences, which shapes behavior by encoding desirable future outcomes. An intrinsic motivation approach called the Maximum Occupancy Principle (MOP) proposes that agents act so as to maximize occupancy over future paths of states and actions, with no preferences or epistemic targets. Despite its simple formulation, MOP gives rise to rich and adaptive behaviors that combine exploratory variability with goal-directed dynamics. In this work, we extend MOP to partially observable environments and introduce a Bellman reformulation of the Expected Free Energy for Active Inference, both incorporating belief-based inference over hidden states as part of the agent state. The Bellman formulation enables tractable offline computation via value iteration over the full belief-state space. We compare the resulting behaviors in a set of minimal experimental settings with uncertain food sources. We find that MOP agents switch between goal-directed (food seeking) behavior and exploration between different food sources, depending on their energy available and their belief state. In contrast, Active Inference agents mostly inhabit regions around a single food source, a strategy having both high pragmatic and epistemic value. We finally compare with Empowerment, which is shown to be qualitatively similar to Active Inference.

Inference on paths of yeast transcription pre-initiation complex assembly by model-based analysis of altered occupancy profiles
E. A. Novikova, A. Goldar, P. Nicolas +1
bioRxiv30 Sep 2026PreprintCC BY-NC-NDFull textBioinformatics & computational biologybioRxiv
The Reuse-and-Append Memory Principle: Application to Latent Cause Inference
Z. Ben Houidi, S. J. Gershman
bioRxiv9 Sep 2026PreprintCC BYFull textNeurosciencebioRxiv
Reframing the lived experience of persistent dizziness: insights from active inference.
Marc A Broberg, Mick Thacker, Karl J Friston +1
Frontiers in psychology14 Oct 2026Peer-reviewedCC BYFull textNeuroscienceEurope PMCPMC
A neuro-inspired computational framework for AGI: predictive coding, active inference and free energy minimisation.
Lioba C S Berndt, Alexander D Shaw
Frontiers in human neuroscience8 Oct 2026Peer-reviewedCC BYFull textNeuroscienceEurope PMC
Two Routes to Interoceptive Prioritisation: Dissociating Likelihood Precision and Preference Reweighting in Active Inference
St John Grimbly, Nicolas Kuske, Evert A. Boonstra +6
arXiv13 Oct 2026PreprintarXiv licenceFull textNeurosciencearXiv

Cite

BibTeX RIS

BibTeX

@misc{mylonas2026beliefbased,
  title = {{Belief-Based Maximum Occupancy Principle and Active Inference}},
  author = {Manolis Mylonas and Rubén Moreno Bote},
  howpublished = {arXiv},
  year = {2026},
  month = sep,
  doi = {10.48550/arxiv.2609.39342},
  eprint = {2609.39342},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2609.39342},
}

RIS

TY  - UNPB
TI  - Belief-Based Maximum Occupancy Principle and Active Inference
AU  - Manolis Mylonas
AU  - Rubén Moreno Bote
PY  - 2026
DA  - 2026/09/30
PB  - arXiv
DO  - 10.48550/arxiv.2609.39342
UR  - https://arxiv.org/abs/2609.39342
AB  - Intrinsic motivation plays a central role in adaptive and goal-directed behavior by conferring agents reward-independent objectives and biases useful to act in noisy and uncertain environments. Active Inference addresses the problem of acting in a partially observable environment through a principled framework for belief updating and action selection. A key component of Active Inference is the specification of prior preferences, which shapes behavior by encoding desirable future outcomes. An intrinsic motivation approach called the Maximum Occupancy Principle (MOP) proposes that agents act so as to maximize occupancy over future paths of states and actions, with no preferences or epistemic targets. Despite its simple formulation, MOP gives rise to rich and adaptive behaviors that combine exploratory variability with goal-directed dynamics. In this work, we extend MOP to partially observable environments and introduce a Bellman reformulation of the Expected Free Energy for Active Inference, both incorporating belief-based inference over hidden states as part of the agent state. The Bellman formulation enables tractable offline computation via value iteration over the full belief-state space. We compare the resulting behaviors in a set of minimal experimental settings with uncertain food sources. We find that MOP agents switch between goal-directed (food seeking) behavior and exploration between different food sources, depending on their energy available and their belief state. In contrast, Active Inference agents mostly inhabit regions around a single food source, a strategy having both high pragmatic and epistemic value. We finally compare with Empowerment, which is shown to be qualitatively similar to Active Inference.
ER  -

Abstract

Related

Cite