BB-CBE-3c: Dual-Scale Exceptional-Footprint Test

This project replaces the saturated footprint detector with a preregistered dual-scale detector and evaluates the unpolarized Gowdy production data through the sealed τ ≤ 18 window.

Result

CBE-D-architecture — sealed-window nonclosure; durable dormancy.

• The dual-scale detector passes its narrow-spike, unseen-spike, and nonshrinking-feature controls.
• In production, the exceptional feature is absent through τ=14, appears by τ=16, and is still growing at τ=18.
• The ordinary mask retains 86.4% of the domain, so the earlier saturation defect is resolved.
• The available history ends before the feature turns over, preventing a zero-measure or ordinary-sector conclusion.

Contents

• run_cbe3c.py — original dual-scale campaign.
• run_cbe3c_r1.py — final control-qualified revision.
• CBE3C_RESULTS.json, CBE3C_R1_RESULTS.json — formal records.
• cbe3c*_log.txt — execution logs.

Reproduction

Clone bb-cbe3, bb-cbe3t, and bb-cbe3b beside this repository. Requires Python 3, NumPy, and SciPy.

python3 run_cbe3c_r1.py

Scope

The sealed history is too short to decide almost-everywhere equivalence. Reopening requires a separately preregistered deeper-history campaign without another detector redesign.