Philotechnology / GLCE V38 — Reproducibility and Audit Archive

Philotechnology / GLCE V38 — Reproducibility and Audit Archive Author: Pedro Cervelleira de MelloRelease date: November 1, 2026Resource type: Dataset / Computational Research Archive Overview This repository provides the computational evidence, experimental datasets, source code, statistical analyses, and methodological documentation supporting the development of Philotechnology, a six-stage epistemological framework proposed by Pedro Cervelleira de Mello for transforming philosophical propositions into operational mathematical models, falsifiable predictions, and auditable scientific decisions. The archive contains 184 files, documenting the retrospective development and evaluation of the General Law of Cosmic Evolution (GLCE) through successive computational formulations, culminating in version V38. The research investigates whether philosophical distinctions can motivate mathematically explicit operators whose consequences can be evaluated against independent reference models, observational data, and established computational alternatives. Archive Contents The repository includes: Historical source code and computational records from successive GLCE formulations. Synthetic cosmological simulations involving 13,824 nodes and 20 historical random seeds. Matched-budget comparisons using 30,000-edge networks. Statistical comparisons of V36, V37, and V38, including ablation studies, confidence intervals, and negative findings. Reconstruction of the sparse fixed-network baseline, including 600,000 recorded edge assignments. Observational data comprising 850 terminal networks of Physarum polycephalum across 86 experimental configurations. Frozen predictions generated by a V38-inspired biological adapter and a Tero-family comparator. Portable reproduction scripts, SHA-256 integrity manifests, provenance documentation, and methodological PASS/FAIL assessments. Principal Experimental Findings Synthetic Cosmology GLCE V38 achieved a mean spatial Jaccard index of 71.03%, compared with 68.88% for V36 and 65.14% for the matched-budget sparse fixed-network baseline. The V38 advantage over the sparse fixed control was +5.89 percentage points. These comparisons are exploratory because the same 20 simulation seeds were examined during model development. Furthermore, a historical dense fixed-network comparison achieved 79.68% Jaccard under a different edge budget, preventing unrestricted claims of adaptive superiority. The cosmological reference is computationally generated and does not constitute direct astronomical observation. Observed Biological Networks Across 86 experimental configurations and 850 terminal Physarum networks, the frozen V38-inspired adapter achieved 54.84% mean edge Jaccard, compared with 58.82% for the Tero-family implementation and 56.19% for the initial geometric network. The adapter therefore underperformed both the established comparator and the unlearned geometric baseline. This negative result is explicitly preserved as evidence against the predictive superiority of the tested biological adapter. The experiment does not constitute a literal evaluation of the complete original V38 physical formulation, because essential specimen-specific physical variables and initial conditions were unavailable. Reproducibility and Methodological Limitations This archive preserves historical computational artifacts alongside explicitly identified retrospective reconstructions, derived documentation, and reproduction utilities. Selected cosmological simulations and the biological prediction pipeline have been computationally reproduced using archived reference data. Complete independent replication, regeneration of the underlying cosmological reference simulations, and prospective validation remain open. The experimental findings neither establish nor refute the universal validity of GLCE. Similarly, the proposed Philotechnology framework has not yet demonstrated generalized methodological superiority over existing scientific modeling practices. Scientific and Epistemological Contribution The principal contribution of this deposit is an auditable historical record of the transition from philosophical intuition to mathematical operationalization, computational experimentation, empirical comparison, and explicit methodological decisions. Philotechnology formalizes a six-stage investigative procedure encompassing conceptual distinctions, operational constraints, mathematical operators, falsifiable predictions, comparative testing, and decisions to retain, revise, reject, or suspend particular formulations. The procedure itself remains open to criticism, empirical evaluation, and methodological revision. The archive is intended to facilitate independent scrutiny, reproducibility, comparison with alternative modeling approaches, and future prospective investigations. Original Dataset Attribution and Licensing The observational Physarum polycephalum data originate from: Sakiyev, Timur. PhysarumAI Dataset: Experimental Networks of Physarum polycephalum, Version 3.0 (2026). Zenodo DOI: https://doi.org/10.5281/zenodo.23075204 The original dataset is distributed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. Original observational data and derived extracts are included with attribution to the dataset author. The licensing of independently authored research materials is documented separately in the archive. Historical filenames retaining the acronym LGEC are preserved for computational provenance and reproducibility. GLCE (General Law of Cosmic Evolution) is the current English designation. Final Statement This archive documents a research process in which mathematical models motivated by a philosophical ontology were subjected to computational testing, comparative evaluation, and explicit acknowledgment of negative results. Its purpose is not to claim final validation of a universal law, but to preserve the evidence necessary for further investigation, criticism, and methodological development.