Mathematical Bridge Between IIT and FEP Proposed via MaxCal
A new study on arXiv has uncovered a detailed mathematical relationship between Integrated Information Theory (IIT) and the Free Energy Principle (FEP). The researchers define information as the difference ψ between actual dynamics and a constrained maximum-caliber (MaxCal) path ensemble over a given period. They demonstrate that the cause/effect frameworks of IIT 3.0 stem directly from MaxCal variational principles, allowing for a new derivation of IIT's phenomenological calculus via constrained entropy-maximization (CMEP). This groundbreaking approach presents a theoretical connection to active inference, potentially bridging two key frameworks in understanding consciousness and self-organization.
Key facts
- Paper ID: arXiv:2605.12536v1
- Announce Type: cross
- FEP is a framework for modeling self-organization and learning
- IIT is a computational ontology of consciousness
- Previous unifications lacked rigorous mathematical mapping
- Information defined as deviation from MaxCal path ensemble
- IIT 3.0 cause/effect repertoires derived from MaxCal
- Framework bridges to active inference
Entities
Institutions
- arXiv