Post-Complex Systems I: Cross-Scale Verification from Percolation Phase Transitions to Cognitive Architectures

Fuente: Zenodo
Salvato in:
Dettagli Bibliografici
Autori principali: zhou, changzheng, zhou, ziqing
Natura: Recurso digital
Pubblicazione: Zenodo 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866901210411302912
author zhou, changzheng
zhou, ziqing
author_facet zhou, changzheng
zhou, ziqing
contents <p>The constraint network theory of complex adaptive systems has proposed a<br>series of testable theoretical predictions regarding core mechanisms such as per<br>colation phase transitions, meta-landscape dimensional evolution, recursive depth<br>truncation, renormalization group scaling, and hierarchical cognitive architectures.<br>However, these predictions still lack systematic numerical experimental support<br>on a unified, controllable platform. In this paper, we construct the CAS-Net uni<br>fied computational experimental platform, which takes the weighted directed con<br>straint network as its sole data structure kernel and carries five scenario types—<br>percolation measurement, adaptive walk, hierarchical agent negotiation, network<br>coarse-graining, and multi-agent cognitive tasks—through a modular plug-in archi<br>tecture. The experiments cover: measuring percolation critical exponents on ran<br>dom geometric graphs and Barabási–Albert scale-free networks and testing their<br>universal class membership; observing the emergence of the optimal dimensional<br>ity of the meta-landscape in multi-niche signal environments; statistically analyz<br>ing the recursive depth distribution and verifying structural closure in multi-layer<br>rule agent systems; testing scale invariance under weighted directed modularity<br>guided coarse-graining iterations; and recording the modality-switching dynamics<br>and maintenance of group diversity of agents with a three-layer cognitive archi<br>tecture in non-stationary classification tasks. All experimental results are in good<br>agreement with the theoretical predictions, and the cross-scale consistency check<br>further reveals that the final states of different scenarios spontaneously cluster in a<br>band-shaped critical region centered at the percolation critical point in the connec<br>tion density–cognitive cost phase space. This paper thus provides a reproducible<br>and deductive numerical foundation for the rigorous study of constraint network<br>dynamics.</p>
format Recurso digital
id zenodo_https___doi_org_10_5281_zenodo_19937411
institution Zenodo
language
publishDate 2026
publisher Zenodo
record_format zenodo
spellingShingle Post-Complex Systems I: Cross-Scale Verification from Percolation Phase Transitions to Cognitive Architectures
zhou, changzheng
zhou, ziqing
constraint networks; numerical simulation; percolation phase transition; meta-landscape; cognitive architecture; multi-agent system; scaling law
<p>The constraint network theory of complex adaptive systems has proposed a<br>series of testable theoretical predictions regarding core mechanisms such as per<br>colation phase transitions, meta-landscape dimensional evolution, recursive depth<br>truncation, renormalization group scaling, and hierarchical cognitive architectures.<br>However, these predictions still lack systematic numerical experimental support<br>on a unified, controllable platform. In this paper, we construct the CAS-Net uni<br>fied computational experimental platform, which takes the weighted directed con<br>straint network as its sole data structure kernel and carries five scenario types—<br>percolation measurement, adaptive walk, hierarchical agent negotiation, network<br>coarse-graining, and multi-agent cognitive tasks—through a modular plug-in archi<br>tecture. The experiments cover: measuring percolation critical exponents on ran<br>dom geometric graphs and Barabási–Albert scale-free networks and testing their<br>universal class membership; observing the emergence of the optimal dimensional<br>ity of the meta-landscape in multi-niche signal environments; statistically analyz<br>ing the recursive depth distribution and verifying structural closure in multi-layer<br>rule agent systems; testing scale invariance under weighted directed modularity<br>guided coarse-graining iterations; and recording the modality-switching dynamics<br>and maintenance of group diversity of agents with a three-layer cognitive archi<br>tecture in non-stationary classification tasks. All experimental results are in good<br>agreement with the theoretical predictions, and the cross-scale consistency check<br>further reveals that the final states of different scenarios spontaneously cluster in a<br>band-shaped critical region centered at the percolation critical point in the connec<br>tion density–cognitive cost phase space. This paper thus provides a reproducible<br>and deductive numerical foundation for the rigorous study of constraint network<br>dynamics.</p>
title Post-Complex Systems I: Cross-Scale Verification from Percolation Phase Transitions to Cognitive Architectures
topic constraint networks; numerical simulation; percolation phase transition; meta-landscape; cognitive architecture; multi-agent system; scaling law
url https://doi.org/10.5281/zenodo.19937411