Work

ALP-ttbar Toy MC v2.3 — Post-Registration Results: Coverage Ratio, Hybrid Barrel–Endcap Strategy, and η Scan

Year: 2026
Type: preprint
Abstract: Post-registration results for the ALP-ttbar Toy MC v2.3 preregistered simulation study (Coverage–Hybrid–Eta Scan). This toy Monte Carlo study investigates displaced light ALP (axion-like particle) → e⁺e⁻ decays in ttbar events, focusing on coverage-based material-veto strategies for background rejection. The simulation employs numpy-based mini-tracking with helical propagation, Highland multiple scattering, and Bethe–Heitler bremsstrahlung, combined with a two-hypothesis vertex fit using transverse and longitudinal Δχ² separation. Version 2.3 introduces a pseudorapidity (η) scan across five η points (1.5, 1.7, 2.0, 2.3, 2.5), examining how cosh(η) compression and oblique incidence factors affect material-veto coverage and signal discrimination. Four veto strategies are compared: uniform hard-veto, uniform tag, hybrid (barrel-tag + endcap-veto), and reverse-hybrid, with a combined limit metric defined as 1/μ_combined = F_barrel/μ_barrel + F_endcap/μ_endcap (F_endcap = 0.20). Key results:— Barrel coverage ~34–52% (n_σ = 2–4, σ = 3 mm); endcap coverage ~4.6–6.2% (n_σ = 2–4, σ = 1 mm); barrel-to-endcap ratio ~8:1.— 7 of 8 preregistered predictions confirmed. P1 (strict non-increasing HV degradation with η) refuted: pre-rounding values show small increases between adjacent η points (e.g. 1.058201 → 1.059530 at η = 1.5 → 1.7), violating the strict non-increasing condition. Whether these fluctuations reflect genuine η dependence or stochastic variation from the single-seed simulation is not assessed.— Tag/HV advantage ratio increases monotonically from ~1.21 (η = 1.5) to ~2.78 (η = 2.5), confirming that the tagging strategy's relative benefit grows at forward pseudorapidity.— The hybrid strategy outperforms all uniform strategies at the 17 MeV ALP mass point but not at 100 MeV, consistent with preregistered expectations. Transparency: predictions P2–P5 are consistency/combination checks derivable from previously published v2.1/v2.2 values; predictions P1 and P6–P8 are genuinely new η-scan results. This package includes full Python source code (run_v23.py, coverage.py, mtag_v21.py, mtag_v22.py, endcap_sim.py), numerical results (v23_results.npy), the PDF generation script (make_report_pdf_v23.py), and a 28-figure PDF report. Note: Gate 1 of run_v23.py requires the v2.1 preregistration and v2.2 results directories (../v21_prereg/ and ../v22_results/) to reproduce the full pipeline. Preregistration (frozen): DOI 10.5281/zenodo.24244893GitHub repository: https://github.com/wspaik5/ALP-ttbar-Toy-MC-v2.3-Post-Registration-Results-Coverage-Hybrid-EtaMotivated by: arXiv:2610.38932
Language: English
Open Access status: green
OpenAlex ID: W7233765618
Cites:
Cited by: 2
Related to: 10