Connecting the free energy principle with quantum cognition

Area of Science:

• Cognitive Science
• Theoretical Physics
• Computational Neuroscience

Background:

• The free energy minimization principle, rooted in Bayesian inference, typically employs classical logic (Boolean lattice).
• Quantum cognition, however, utilizes quantum logic (orthomodular lattice) and allows for deviations from experience-based views.
• A perceived conflict exists between these frameworks, hindering unified understanding.

Purpose of the Study:

• To reconcile the free energy minimization principle with quantum cognition.
• To introduce a novel framework, "excess Bayesian inference," that bridges these two fields.
• To demonstrate how this new inference method generates an orthomodular lattice structure.

Main Methods:

• Extending the core concepts of Bayesian inference beyond classical applications.
• Enhancing active inference and embodied intelligence principles.
• Utilizing rough-set lattice techniques on joint probability distributions to derive logical structures.

Main Results:

• "Excess Bayesian inference" is introduced as a bridge between free energy minimization and quantum cognition.
• This novel inference method demonstrably entails an orthomodular lattice, aligning with quantum cognition.
• Classical Bayesian inference inherently entails a Boolean lattice, consistent with classical logic.

Conclusions:

• The proposed "excess Bayesian inference" successfully connects the free energy minimization principle with quantum cognition.
• This framework resolves the apparent conflict by showing how deviations from classical Bayesian inference lead to quantum logical structures.
• The findings suggest a more unified theoretical basis for understanding cognition and decision-making across classical and quantum paradigms.