ALP-ttbar Toy MC v2.5 — Post-Registration Results: Resolution Deconflation (SENSOR_DISK_Z Separation)

Post-registration results for the ALP-ttbar Toy MC v2.5 study: Resolution Deconflation via SENSOR_DISK_Z Separation in Displaced Light ALP → e⁺e⁻ in ttbar Events. The v2.4 endcap simulation used a single DISK_Z array (17 band positions) for both conversion generation and resolution assignment, conflating service-band positions with sensor-disk positions. This meant vertices near passive service material received pixel-quality resolution (σ_z = 1.0 mm) instead of strip-quality (σ_z = 3.0 mm). The v2.5 fix introduces a separate SENSOR_DISK_Z array containing only the 9 sensor disk positions [230, 300, 450, 620, 820, 1100, 1500, 1900, 2300] mm, used exclusively for resolution assignment in _measure_endcap(). The full DISK_Z (17 bands) is retained for conversion generation, veto placement, tag formula evaluation, and nearest-disk pull. v2.5.1 bug fix: The initial run contained a Python import-shadowing bug where sys.path.insert(0, ...) placed the v2.2 endcap_sim.py (without SENSOR_DISK_Z) ahead of the v2.5 version. All reported results are from the corrected v2.5.1 run. Key findings:(1) The resolution deconflation has a significant quantitative effect. HV degradation increases by +0.024× (200 MeV) to +0.047× (50 MeV) across all masses, representing a 42–65% correction relative to the v2.4 HV signal. The conflated resolution was partially compensating for the material effect, causing v2.4 to underestimate the true performance cost of service material.(2) 9 of 10 predictions confirmed using the published v2.3 baseline (P10 refuted: |ΔHV| > |ΔTag|, opposite of predicted). If the recomputed baseline is used, 8 of 10 (P7 and P10 refuted).(3) P3 flips from REFUTED in v2.4 to CONFIRMED in v2.5: the tag/HV ratio decreases by 3.1% from the v2.3 baseline, confirming that the conflation was disproportionately boosting the tag formula's discrimination. Mechanism: ΔHV = +0.045 >> ΔTag = +0.007, raising the denominator of Tag/HV.(4) Validation V3 (200 MeV convergence) is refuted: HV = 1.060× and Tag = 1.061× both exceed the 1.05 threshold.(5) ΔHV is remarkably uniform across pseudorapidity η (+0.044 to +0.047), while ΔTag increases sharply with η (+0.007 at η = 2.0 to +0.110 at η = 2.5). Two versions of the v2.5 preregistration README exist: the DOI-registered version (SHA-256: 4f7b5160...) and an earlier draft (SHA-256 beginning e0d40...). The DOI-registered version is the official preregistration referenced throughout. All 5 sanity gates passed. 5 of 6 validations passed (V3 refuted). Companion to preregistration doi:10.5281/zenodo.24329044. Builds on v2.4 preregistration and results (doi:10.5281/zenodo.24296796, doi:10.5281/zenodo.24302619). Motivated by arXiv:2610.38932 (Niedziela, Alimena et al.). ZIP contains 16 files: README.md, simulation code (run_v25.py, endcap_sim.py, mtag_v24.py, coverage.py, mtag_v21.py, mtag_v22.py), results data (v25_results.npy), chart generator (make_charts.py), PDF generator (make_report_pdf_v25_results.py), 5 comparison charts (PNG), and summary PDF. Independent personal toy study. Not affiliated with the authors of arXiv:2610.38932 or arXiv:2306.08686, nor with the CMS, ATLAS, LHCb or ALICE Collaborations or the FCC study.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-19
DOI
https://doi.org/10.5281/zenodo.24341049
Primary Topic
Particle Detector Development and Performance
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

ALP-ttbar Toy MC v2.5 — Post-Registration Results: Resolution Deconflation (SENSOR_DISK_Z Separation)

Won Shik Paik
10 links
Zenodo (CERN European Organization for Nuclear Research)
Particle Detector Development and Performance
preprint

ALP-ttbar Toy MC v2.5 — Post-Registration Results: Resolution Deconflation (SENSOR_DISK_Z Separation)

Won Shik Paik
preprint en

Abstract

Post-registration results for the ALP-ttbar Toy MC v2.5 study: Resolution Deconflation via SENSOR_DISK_Z Separation in Displaced Light ALP → e⁺e⁻ in ttbar Events. The v2.4 endcap simulation used a single DISK_Z array (17 band positions) for both conversion generation and resolution assignment, conflating service-band positions with sensor-disk positions. This meant vertices near passive service material received pixel-quality resolution (σ_z = 1.0 mm) instead of strip-quality (σ_z = 3.0 mm). The v2.5 fix introduces a separate SENSOR_DISK_Z array containing only the 9 sensor disk positions [230, 300, 450, 620, 820, 1100, 1500, 1900, 2300] mm, used exclusively for resolution assignment in _measure_endcap(). The full DISK_Z (17 bands) is retained for conversion generation, veto placement, tag formula evaluation, and nearest-disk pull. v2.5.1 bug fix: The initial run contained a Python import-shadowing bug where sys.path.insert(0, ...) placed the v2.2 endcap_sim.py (without SENSOR_DISK_Z) ahead of the v2.5 version. All reported results are from the corrected v2.5.1 run. Key findings:(1) The resolution deconflation has a significant quantitative effect. HV degradation increases by +0.024× (200 MeV) to +0.047× (50 MeV) across all masses, representing a 42–65% correction relative to the v2.4 HV signal. The conflated resolution was partially compensating for the material effect, causing v2.4 to underestimate the true performance cost of service material.(2) 9 of 10 predictions confirmed using the published v2.3 baseline (P10 refuted: |ΔHV| > |ΔTag|, opposite of predicted). If the recomputed baseline is used, 8 of 10 (P7 and P10 refuted).(3) P3 flips from REFUTED in v2.4 to CONFIRMED in v2.5: the tag/HV ratio decreases by 3.1% from the v2.3 baseline, confirming that the conflation was disproportionately boosting the tag formula's discrimination. Mechanism: ΔHV = +0.045 >> ΔTag = +0.007, raising the denominator of Tag/HV.(4) Validation V3 (200 MeV convergence) is refuted: HV = 1.060× and Tag = 1.061× both exceed the 1.05 threshold.(5) ΔHV is remarkably uniform across pseudorapidity η (+0.044 to +0.047), while ΔTag increases sharply with η (+0.007 at η = 2.0 to +0.110 at η = 2.5). Two versions of the v2.5 preregistration README exist: the DOI-registered version (SHA-256: 4f7b5160...) and an earlier draft (SHA-256 beginning e0d40...). The DOI-registered version is the official preregistration referenced throughout. All 5 sanity gates passed. 5 of 6 validations passed (V3 refuted). Companion to preregistration doi:10.5281/zenodo.24329044. Builds on v2.4 preregistration and results (doi:10.5281/zenodo.24296796, doi:10.5281/zenodo.24302619). Motivated by arXiv:2610.38932 (Niedziela, Alimena et al.). ZIP contains 16 files: README.md, simulation code (run_v25.py, endcap_sim.py, mtag_v24.py, coverage.py, mtag_v21.py, mtag_v22.py), results data (v25_results.npy), chart generator (make_charts.py), PDF generator (make_report_pdf_v25_results.py), 5 comparison charts (PNG), and summary PDF. Independent personal toy study. Not affiliated with the authors of arXiv:2610.38932 or arXiv:2306.08686, nor with the CMS, ATLAS, LHCb or ALICE Collaborations or the FCC study.

Zenodo (CERN European Organization for Nuclear Research)
Particle Detector Development and Performance
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon