Functional Modeling of Neural Topology

Fuente: Zenodo
Gespeichert in:
Bibliographische Detailangaben
1. Verfasser: Williams, Andy E
Format: Recurso digital
Veröffentlicht: Zenodo 2026
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866901666819735552
author Williams, Andy E
author_facet Williams, Andy E
contents <p>Cognitive differences across individuals and species are typically characterized through trait-based measurements---IQ scores, symptom inventories, or isolated functional capacities---that lack mechanistic continuity across development, evolution, and pathology. This paper proposes an alternative: cognition as adaptive network topology under feedback. We define a layered functional model of cognitive network growth in which the system is decomposed into three non-collapsible layers: a Governance Layer (active, multi-attractor, non-tree-like), a Growth Layer (tree-like under scalarized fitness), and a Topology Layer (the resulting network structure). Neurons serve as stable, addressable functional units; cognitive transitions correspond to directed edges with associated costs and strengths; and reasoning styles emerge from the interaction of operator families with governance meta-controllers. When governance temporarily collapses fitness dimensionality onto a scalar objective, growth dynamics follow tree-like adaptive laws, producing qualitatively distinct topologies: dense, locally-branched networks versus sparse, trunk-dominant architectures with long-range bridges. The framework explains evolutionary phase transitions, particularly the emergence of language as a bridging operator that dramatically reduces cross-domain construction costs. All claims are framed as falsifiable predictions about network morphology rather than clinical assertions, with explicit specification of the domain of validity for each modeling component.</p>
format Recurso digital
id zenodo_https___doi_org_10_5281_zenodo_18360817
institution Zenodo
language
publishDate 2026
publisher Zenodo
record_format zenodo
spellingShingle Functional Modeling of Neural Topology
Williams, Andy E
<p>Cognitive differences across individuals and species are typically characterized through trait-based measurements---IQ scores, symptom inventories, or isolated functional capacities---that lack mechanistic continuity across development, evolution, and pathology. This paper proposes an alternative: cognition as adaptive network topology under feedback. We define a layered functional model of cognitive network growth in which the system is decomposed into three non-collapsible layers: a Governance Layer (active, multi-attractor, non-tree-like), a Growth Layer (tree-like under scalarized fitness), and a Topology Layer (the resulting network structure). Neurons serve as stable, addressable functional units; cognitive transitions correspond to directed edges with associated costs and strengths; and reasoning styles emerge from the interaction of operator families with governance meta-controllers. When governance temporarily collapses fitness dimensionality onto a scalar objective, growth dynamics follow tree-like adaptive laws, producing qualitatively distinct topologies: dense, locally-branched networks versus sparse, trunk-dominant architectures with long-range bridges. The framework explains evolutionary phase transitions, particularly the emergence of language as a bridging operator that dramatically reduces cross-domain construction costs. All claims are framed as falsifiable predictions about network morphology rather than clinical assertions, with explicit specification of the domain of validity for each modeling component.</p>
title Functional Modeling of Neural Topology
url https://doi.org/10.5281/zenodo.18360817