What question did this study set out to answer?

This work aims to compile and analyze the findings of the GCFT dark-sector book arc while improving reliability and transparency.

June 18, 2026Open Access

CDN-250 · The Dark-Sector Arc — GCFT Cluster Book Arc, Collected Edition (Art Edition)

Key Points

This work aims to compile and analyze the findings of the GCFT dark-sector book arc while improving reliability and transparency.
Establish a seven-term κ-proxy analyzed across 25 galaxy clusters from various surveys.
Implement a blind holdout protocol to ensure the integrity of the test design and data handling.
Audit and readjust prediction bands based on empirical percentile distributions assessed from prior analyses.
Achieved 98.97% mean dark-sector closure with a 0.215% population non-GCFT residual.
Identified an irreducible P-class bound at ≤0.29% for non-GCFT fraction across the tested clusters.
Diagnosed a CONFLICT in band specifications, leading to a successful re-evaluation and full PASS on the holdout set.

Abstract

CDN-250 is the collected edition of the GCFT dark-sector book arc (CDN-227 through CDN-249), narrating the complete journey from the parity-split framework through holdout PASS in a single document. No new computation is performed; all results are traceable to predecessor Zenodo DOIs. The arc establishes that a seven-term κ-proxy, fitted on 25 galaxy clusters from CLASH/CCCP/WtG surveys (CDN-240), reduces S-class residual mass by 43% (CDN-242 Track 1), identifies an irreducible P-class bound at ≤0. 29% population-averaged non-GCFT fraction (CDN-242 Track 3), and discovers a genuine F-class super-ceiling population (CDN-242 Track 2). A blind holdout protocol pre-registers the test design before revealing data (CDN-243), catches and discards a contaminated 15-cluster sample (CDN-244/245), and rebuilds with an audit gatekeeper (CDN-246). The holdout execution (CDN-247) finds 98. 97% mean dark-sector closure and 0. 215% population non-GCFT residual — comfortably below the 0. 29% bound — with zero single-cluster cap violations, but trips a CONFLICT verdict on aspirationally-set per-cluster prediction bands (10/15 in-band, below the 80% PASS threshold). CDN-248 diagnoses the CONFLICT as a band-specification artefact: the S-band lower edge (0. 10) exceeds 83% of CDN-242's own S-class and 100% of the holdout S-class. CDN-249 re-anchors the bands to empirical percentile distributions and re-scores the holdout as 15/15 PASS with no proxy refitting. Standing claims: 0. 215% population non-GCFT bound on independent holdout; 98. 97% mean closure; E/S/P taxonomy reproduces out-of-sample; self-correction protocol works as designed. Open threads: F-class super-ceiling unaddressed (0/2 tested, declared limitation) ; literature-vector replacement deferred; band generalization to other surveys (SPT-SZ, ACT) untested. Forbidden claims stated explicitly: GCFT does not claim to eliminate dark matter; the κ-proxy is phenomenological; F-class is not tested on holdout. Complete DOI chain for CDN-227 through CDN-249 included. Art Edition with enhanced star-field design system (seed 250, gold accent). Lock: CDN250DARKSECTORARCLOCKED.

Read Full Paperexternally

Perguntar à IA

Bookmark

View Full Paper