CR004 $pp \to B \bar{B} \to tW/bZ/bH + X$

In custodial RS and composite-Higgs language, the \(B\) state is a physical fermion resonance, not the KK gluon itself. A direct \(B\bar B\) search therefore constrains \(m_B\) under a specified custodial multiplet and decay pattern; it becomes a constraint on \(M_{\mathrm{KK}}\) only after choosing the model relation between the light bottom partner and the gauge/fermion KK scale. Pair production is comparatively robust because it is QCD driven, but the mass exclusion still depends on the signal width, the allowed decays, and the branching fractions into \(tW\), \(bZ\), and \(bH\). For the RS scan in this repo, CR004 is a cross-check rather than the leading bound. The this repository quark-scan methodology note reports \(M_{\mathrm{KK}}^{\min}(p50,g_\ast=3)=47.26~\mathrm{TeV}\) from the low-energy flavor scan, with \(127.13~\mathrm{TeV}\) at \(95\%\) acceptance. That scale is far above the present direct \(B\)-partner reach near \(1.5~\mathrm{TeV}\). The collider entry is still useful for non-anarchic RS variants, custodial spectra where fermion partners sit well below the gauge KK scale, or benchmark points tuned to evade the low-energy flavor envelope.

The pre-LHC RS-motivated baseline is Contino--Servant arXiv:0801.1679, which emphasized same-sign dileptons from \(B,T_{5/3}\to Wt\) in custodial top-partner spectra. Post-2010 LHC work turned that benchmark strategy into broad branching-fraction searches: ATLAS arXiv:1808.02343 combined 13 TeV searches across the \(T\) and \(B\) decay simplex; CMS arXiv:2008.09835 made fully hadronic \(bH/bZ\) final states competitive with leptonic channels; CMS arXiv:2209.07327 used single-lepton, same-sign dilepton, and multilepton categories and drives the PDG \(B\to Wt\) corner; and ATLAS arXiv:2210.15413 plus arXiv:2212.05263 are full-Run-2 \(Z+\)third-generation and high-\(E_T^{\rm miss}\) cross-checks. The current CMS arXiv:2402.13808 result added a \(0\ell/2\ell\) combination with HEPData exclusion contours over the \(bH\), \(bZ\), and \(tW\) branching simplex. The CMS arXiv:2405.17605 review is used as Run-2 synthesis, not as a new measured limit.

The quoted limits are valid as direct LHC mass-exclusion statements for prompt, narrow enough, pair-produced \(B\) quarks with the stated branching assumptions. They are not inclusive limits on every custodial-RS bottom partner. A broad resonance, nonstandard cascade decays, additional nearly degenerate partners, altered \(B\)-tag/Higgs/Z acceptances, or branching fractions spread across several modes can move the bound. A single production bound can reach higher masses, but it is controlled by electroweak mixing/coupling choices and is therefore not the canonical CR004 pair-production observable.

NO. The required code-coverage grep over quarkConstraints/, qcd/, flavorConstraints/, neutrinos/, yukawa/, warpConfig/, solvers/, scanParams/, and tests/ found no direct collider constraint, VLQ object, branching-simplex filter, or ATLAS/CMS/HEPData limit ingestion. The collider/VLQ query returned only tests/test\_alpha\_s.py:88 and tests/test\_alpha\_s.py:89, a \(\alpha_s\) CMS/RunDec example, not an LHC search. Nearby evidence is flavor-only: scanParams/scan.py:399 computes \(M_{\mathrm{KK}}\) for a scan row and scanParams/scan.py:523 applies the \(\mu\to e\gamma\) check, while quarkConstraints/scan.py:359 maps \(\Lambda_{\rm IR}\) to \(M_{\mathrm{KK}}\) and quarkConstraints/scan.py:377 evaluates \(\Delta F=2\) constraints.

HIGH. A faithful implementation needs external LHC likelihoods or a reinterpretation chain such as CheckMATE, MadAnalysis5, or SModelS, plus signal generation for the custodial spectrum, widths, branching fractions, detector acceptances, and correlations across final states. A simple hard-coded mass veto would be easy but would be misleading for RS scans because \(m_B/M_{\mathrm{KK}}\), widths, and \(B\to tW,bZ,bH\) fractions are model dependent.

PDG2025\_BPrimeListing; ContinoServant2008\_TopPartners; ATLAS2018\_VLQCombination; CMS2020\_BVLQFullyHadronic; CMS2023\_VLQLeptonic; ATLAS2023\_ZThirdGen; ATLAS2023\_METPartners; CMS2024\_BVLQDilepHad; HEPData2024\_BVLQContours; CMS2025\_VLQReview.