ALP-ttbar Toy MC v2.6 — Preregistration: Multi-Seed Ensemble for Statistical Validation of Resolution-Deconflation Results
Year: 2026
Type: preprint
Abstract: Preregistration for ALP-ttbar Toy MC v2.6: Multi-Seed Ensemble in Displaced Light ALP → e⁺e⁻ This preregistration freezes the analysis plan for v2.6, which extends the single-seed v2.5 results (DOI: 10.5281/zenodo.24341050) to a multi-seed ensemble of 50 independent random-seed pairs. The goal is to place confidence intervals on all degradation ratios (HV, Tag, Q) and on the deconflation deltas that isolate the effect of the SENSOR_DISK_Z resolution fix introduced in v2.5. Motivation. Several v2.5 results sit near preregistered decision boundaries, and their confirmation or refutation could depend on stochastic fluctuations inherent to the single pair of random seeds (SEED_CONV = 5, SEED_SIG = 11) used since v2.4. A multi-seed ensemble quantifies this seed-to-seed variance and determines whether the observed trends are robust. Key design choices:- 50-seed primary ensemble (i = 0–49), with SEED_CONV(i) = 100×i + 5, SEED_SIG(i) = 100×i + 11; seed i = 0 reproduces the published v2.5 baseline exactly.- Per-seed deconflation delta = v2.5 config (17-band geometry + SENSOR_DISK_Z for resolution) minus v2.4-equivalent conflated config (same 17-band geometry + DISK_Z for resolution), isolating the resolution-assignment effect within the same geometry.- Three detector configurations: E3 (v2.5, 17-band + SENSOR_DISK_Z), E3_conflated (v2.4-equivalent, 17-band + DISK_Z resolution), and E3_9disk (v2.3-style, 9-disk only).- 100-seed supplementary ensemble available if the 50-seed primary shows insufficient convergence, without overriding primary results.- Gate criteria (G1–G5) define pass/fail conditions before any computation, including exact numerical reproduction of v2.5 at seed i = 0. Propositions under test:- P3 (seed robustness): whether the median ensemble degradation ratio lies within ±0.05× of the single-seed v2.5 value.- P7 (baseline comparison): whether the ensemble-median d values are consistent with the published v2.3 baseline ratio of 1.030×.- V3 (deconflation): whether the ensemble-mean deconflation delta is positive (confirming the SENSOR_DISK_Z fix systematically increases the degradation signal). Contents of the deposit: README.md (the full preregistration document), a PDF rendering of the same, and SHA256SUMS.txt for independent file verification. Series chain: v2.5 prereg & results (10.5281/zenodo.24329044, 10.5281/zenodo.24341050); v2.4 (10.5281/zenodo.24296796, 10.5281/zenodo.24302619); v2.3 (10.5281/zenodo.24244893, 10.5281/zenodo.24273586); v2.2 (10.5281/zenodo.24189990, 10.5281/zenodo.24238224); v2.1 (10.5281/zenodo.24063091, 10.5281/zenodo.24136462); v2.0 overview (10.5281/zenodo.24056097). Motivated by arXiv:2610.38932 (Alimena, Niedziela et al.). Independent personal toy study; not affiliated with the authors of arXiv:2610.38932 or arXiv:2306.08686, nor with any LHC collaboration or the FCC study. Detector numbers are assumed; only qualitative trends and ratios are meaningful.
Author: Paik, Won Shik
Language: English
Subfield: Nuclear and High Energy Physics
Field: Physics and Astronomy
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7236750078
