Reproducibility Package for "Hierarchical Genesis through Learning: From Learned Macro-Representations toward New Organizational Levels"
This record provides the computational reproducibility package for the study “Hierarchical Genesis through Learning: From Learned Macro-Representations toward New Organizational Levels.” Stage A is a 20-seed gradient-learning experiment in which a 24-component adaptive system develops a low-rank macro–micro feedback pathway. Stage B is an independent 25-seed structural qualification test evaluating integration, relative macro autonomy, structure-specific controls, dynamic higher-to-lower intervention effects, and forward–reverse hysteresis. The package includes the recovered simulation code, software-environment records, seed-level numerical outputs, H3/H4 threshold-sensitivity results, critical-demand-parameter (Dc) sensitivity results, and a verification report. The recovered implementation was rerun from scratch and reproduced the archived principal numerical CSV outputs exactly in the verified environment. This package is intended to support transparent verification, reproduction, and reuse of the computational analyses reported in HGT-III.