Gorde:
| Egile nagusia: | |
|---|---|
| Formatua: | Recurso digital |
| Hizkuntza: | |
| Argitaratua: |
Zenodo
2026
|
| Gaiak: | |
| Sarrera elektronikoa: | https://doi.org/10.5281/zenodo.19601789 |
| Etiketak: |
Etiketa erantsi
Etiketarik gabe, Izan zaitez lehena erregistro honi etiketa jartzen!
|
Aurkibidea:
- <p>Standard predictive processing frameworks treat all friction — the resistance an organism encounters when acting on its environment — as prediction error: a signal that the generative model requires revision. This paper argues that this identification is fundamentally incomplete. We propose a formal decomposition of the organism's friction signal into two irreducibly distinct components: a model-correction component (IT_ε) arising from miscalibration of the generative model, and a structural-constraint component (IT_Φ) arising from the organism approaching the boundary of what its environment genuinely permits. The structural-constraint component is formalized through the Exclusion Topology Φ: a graded, probabilistic set of constraints on the organism's state-space that represent external invariants rather than internal modeling failures. Critically, and in response to quantum-mechanical considerations, Φ is not defined as a set of absolutely impossible transitions (probability exactly zero), but as the set of transitions whose probability approaches zero under all admissible physical and informational transformations — a definition robust to quantum tunneling and stochastic fluctuation while preserving the core insight that some constraints are model-independent. We establish the intersubjective invariance of Φ across observers with different generative models, derive a thermodynamic lower bound showing that constraint-encounter friction is energetically irreducible, and show that optimal cognition is precisely the efficient navigation of the constraint boundary rather than its elimination. Four testable predictions are developed. The paper is self-contained: all required constructs are developed from first principles within the Unified Dynamic</p>