Same-Sign Dilepton + Jets Signature

• The same-sign dilepton plus jets signature is a rare event topology defined by two isolated, identically charged leptons, multiple jets, and often significant MET.
• Its precise event selection and low SM background enable clear discrimination of new physics signals, with rigorous data-driven methods and simulations.
• The signature is a versatile probe for scenarios including supersymmetry, Majorana neutrinos, and vector-like quarks, validating constraints through detailed statistical analysis.

The same-sign dilepton plus jets signature is a highly distinctive event topology at hadron colliders, characterized by two isolated leptons of identical electric charge (e.g., $e^{\pm} e^{\pm}$, $\mu^{\pm} \mu^{\pm}$, $e^{\pm} \mu^{\pm}$), multiple hadronic jets, and, in many searches, significant missing transverse energy (MET). Owing to the extreme rarity of such events in the Standard Model (SM)—arising mainly from rare gauge-boson or top-associated processes—this signature provides a powerful probe for numerous scenarios of new physics, including supersymmetry (SUSY), models with heavy Majorana neutrinos, vector-like quarks, extended gauge sectors, and exotic scalar or leptoquark states. The theoretical, phenomenological, and experimental foundations of this signature have been developed in a series of landmark studies, with a particular focus on inclusive and b-tagged final states at the LHC (Collaboration, 2012, Melia, 2011, Lei, 2013, Collaboration, 2012, Alves et al., 2012, Cui et al., 2022, 0909.4300, Collaboration, 2011, Collaboration, 2020, Arganda et al., 2023, Collaboration, 2016, Collaboration, 2013, Collaboration, 2012, Collaboration, 2014, Weinberg, 2011, Berger et al., 2013, Collaboration, 2011).

1. Theoretical Foundations and Motivations

The same-sign dilepton plus jets final state is an "exotic" topology within the SM. The dominant irreducible SM processes are rare: QCD-mediated $pp \to W^\pm W^\pm jj$ production, $t\bar{t}W$, $t\bar{t}Z$, and $WZ$, along with extremely suppressed triboson and double-parton scattering contributions (Melia, 2011, Collaboration, 2012, Lei, 2013). Due to the electrically-charged nature of the W bosons, two like-sign leptons are produced only under strict charge-flow conditions in SM diagrams, often requiring each $W^\pm$ to be radiated off a different quark line. For instance, the QCD-driven $pp \to W^+W^+jj$ process yields a finite $2 \to 4$ phase-space even with looser jet requirements, but the cross section remains extremely small at high energies.

From the BSM perspective, the SSDL+jets signature provides sensitivity to a wide array of theories:

• Supersymmetry: Pair-production of strongly interacting superpartners (gluinos/squarks) followed by cascade decays through Majorana gluinos or charginos can yield SSDL as a hallmark, especially in R-parity-conserving scenarios where significant MET is also present (Collaboration, 2012, Lei, 2013, Collaboration, 2012, 0909.4300, Collaboration, 2014, Berger et al., 2013)
• R-parity violating (RPV) and Minimal Flavor Violation (MFV) SUSY: Same-sign dileptons persist even without large MET, as in cases where gluinos decay to stops and stops decay via $UDD$-type baryon-number violating couplings (Berger et al., 2013).
• Vector-like and Fourth-Generation Quarks: Models with heavy partners such as $T$ (charge $+2/3$) or $B$ quarks, which decay as $T\to Wb$/$Wt$, $B\to Wt$, produce SSDL signatures when both heavy quarks decay semi-leptonically (Cui et al., 2022, Lei, 2013).
• Majorana Neutrinos: Processes involving Majorana states allow lepton-number violating decays yielding SSDL+jets, e.g., $pp\to \ell^\pm N\to \ell^\pm\ell^\pm jj$ (Lei, 2013, Collaboration, 2011).
• Exotic Scalars and Leptoquarks: New gauge multiplets, doubly-charged Higgs bosons, or exotic fermions such as the leptoquark $J_3$ in 3-3-1 models may yield extremely clean SSDL plus jets signals with negligible SM backgrounds (Alves et al., 2012).
• Effective-Operator and Contact-Interactions: 4-quark dimension-6 operators can directly mediate $pp\to tt+jj$ or similar topologies with SSDL, providing a robust path for probing high-scale new physics (Arganda et al., 2023).

Thus, the SSDL+jets signature is particularly valued for its background suppression, model-independence, and high signal-to-background ratio across a broad spectrum of new physics.

2. Experimental Strategy and Event Selection

The canonical event selection for SSDL+jets searches, as implemented by CMS and ATLAS, requires exactly two well-identified, isolated leptons of the same electric charge (electron, muon, or hadronic $\tau$), two or more jets reconstructed with anti-$k_T$ algorithms (radius $R=0.4$–0.5), and various kinematic thresholds on $p_T$, $\eta$, and separation $\Delta R$ (Collaboration, 2012, Collaboration, 2012, Collaboration, 2013, Collaboration, 2014). Further refinements target the b-jet content (for scenarios involving third-generation squark or heavy quark decays), MET, scalar $H_T$ (sum of jet $p_T$), and additional variables such as minimum lepton $p_T$ and invariant mass.

A representative set of criteria (CMS, $\sqrt{s}=7$ TeV, 4.98 fb$^{-1}$) (Collaboration, 2012):

• Exactly two same-sign leptons (e, $\mu$, $\tau_h$), no third lepton forming a $Z$ boson
• Leptons: $|\eta| < 2.4$, from common primary vertex; isolation sum $p_T({\rm tracks}) + E_T({\rm calo})$ in $\Delta R < 0.3$ must be $< 0.15\,p_T$ (lepton)
• Jets: anti-$k_T$ on particle flow, $p_T > 40$ GeV, $|\eta| < 2.5$
• MET reconstructed from particle flow: $MET = |\sum_i \vec{p}_{Ti}|$
• $H_T$ defined as scalar sum of jet $p_T$ (excluding those too close to leptons)
• Invariant mass of dilepton $> 8$ GeV

Signal regions are then defined by various thresholds and binnings in $H_T$, $MET$, and counts of jets and b-tagged jets; for example, five regions in the ($H_T$, $MET$) plane were used to cover a variety of new-physics spectra (see table below).

Region	$H_T$ (GeV)	$MET$ (GeV)
1	$>80$	$>120$
2	$>200$	$>120$
3	$>450$	$>50$
4	$>450$	$>120$
5	$>450$	$>0$

Variation in selection thresholds and object definitions facilitates the coverage of spectra ranging from R-parity-conserving SUSY (large $MET$, hard $H_T$) to RPV or heavy vector-like quark scenarios (hard jets, little MET).

3. Background Estimation Methodologies

Three principal SM backgrounds are systematically treated in all SSDL+jets searches (Collaboration, 2012, Melia, 2011, Lei, 2013, Collaboration, 2012, Collaboration, 2014, Collaboration, 2013):

A. Prompt Same-Sign Dilepton Production (Irreducible):

This category includes SM processes genuinely producing SSDL:

• $t\bar{t}W$, $t\bar{t}Z$, and $W^\pm W^\pm jj$ constitute the dominant irreducible component ($\sim$95%). Their rates are obtained from Monte Carlo simulation (e.g., MadGraph + Pythia), normalized to NLO cross sections; a 50% systematic uncertainty is commonly assigned to cover theoretical predictions.

B. Nonprompt ("Fake") Leptons:

Instruments- or physics-induced misidentification, including:

• Heavy-flavor decays, hadrons misidentified as leptons, jets faking $\tau$.
• Estimated using "tight-to-loose" tag-and-probe in dijet or jet-enriched control samples. The TL ratio (probability for a lepton failing tight cuts to pass them) is measured as a function of $p_T$ and $\eta$ and then applied to data in relevant sidebands. This component often constitutes 20–60% of the total background. A systematic uncertainty of up to 50% is routinely assigned.

C. Charge Misidentification:

Primarily affects electrons via hard bremsstrahlung and subsequent charge assignment errors. The charge-flip rate is measured in $Z\to ee$ and/or $Z\to\tau\tau$ events and is less than 5% of the total background. For muons, the charge misidentification rate is negligible ($\sim 10^{-5}$).

A representative background composition is:

Background Type	Fraction (%)
Irreducible SM	30–70
Nonprompt (fake)	20–60
Charge mis-ID	$<$5

All CMS and ATLAS analyses utilize data-driven control regions, tight-loose matrix methods, and auxiliary MC predictions with NLO normalization. Control regions are validated and uncertainties propagated in statistical inference.

4. Signal Efficiency, Acceptance, and Recasting

The signal efficiency ($\epsilon$) and overall acceptance ($A$) are measured using a combination of control data (\textit{e.g.}, $Z\to\ell\ell$) and fully simulated MC. Scale factors are applied for data-MC discrepancies. The overall approach enables straightforward theoretical recasting and reinterpretation across models (Collaboration, 2012, Collaboration, 2012, Collaboration, 2013).

Electron and muon reconstruction efficiencies rapidly approach plateaus as a function of $p_T$:

• Electron: $\epsilon_e(p_T) \simeq 0.94$ for $p_T > 20$ GeV, with uncertainty 3%.
• Muon: $\epsilon_\mu(p_T) \simeq 0.98$ for $p_T > 20$ GeV, similar uncertainties.
• Hadronic $\tau$: plateau at 34% with 10% uncertainty.

For generator-to-reconstruction emulations, error-function-based parameterizations provide practical mapping. For example (Collaboration, 2012):

$$\epsilon_e(p_T)=0.72\,\mathrm{erf}[(p_T-10)/22.5]+0.22\,[1-\mathrm{erf}((p_T-10)/22.5)]$$

Similarly, $H_T$ and $MET$ thresholds are parametrized with error functions (see Section 4 of (Collaboration, 2012)).

For recasting, users can compute generator-level lepton $p_T$, jet $H_T$, and neutrino+LSP $MET$ and apply the published efficiency parameterizations to obtain expected yields for arbitrary NP spectra. The analytic framework is robust to approximately 15% when compared to full-simulation acceptances.

5. Statistical Analysis and Limits

Limits are set using the modified frequentist $\mathrm{CL}_s$ approach (Collaboration, 2012, Collaboration, 2013, Collaboration, 2012, Collaboration, 2014), which incorporates Poisson statistics for event counts, together with log-normal or Gaussian priors for systematic uncertainties. The observed number of signal ($N_\mathrm{obs}$) and background ($N_\mathrm{bkg}$) events, with uncertainties, are used to construct a likelihood:

$$L(N_\mathrm{obs}\,|\,N_\mathrm{bkg}+N_\mathrm{sig})$$

with marginalization over nuisance parameters reflecting the systematic uncertainties (14–20% on total signal yield depending on the region in (Collaboration, 2012)). The $95\%$ CL exclusion is set by requiring $\mathrm{CL}_s\leq0.05$, leading to limits on the signal cross section:

$$\sigma_{\mathrm{UL}} = \frac{N_{\mathrm{UL}}}{\mathcal{L}\,A\epsilon}$$

where $N_{\mathrm{UL}}$ is the observed upper limit on the number of signal events, $\mathcal{L}$ is the integrated luminosity, and $A\,\epsilon$ is the total acceptance and efficiency.

Observed yields agree with SM predictions in all major analyses. For example, in high-$p_T$ dilepton channels (region 4; $H_T>450$ GeV, $MET>120$ GeV) (Collaboration, 2012):

• Predicted background: $4.9 \pm 2.6$, observed: $4$, UL($N_\mathrm{sig}$): $6.2$ events.

6. Results, Constraints, and Phenomenological Impact

Stringent 95% CL upper limits on new-physics cross sections have been set without observing significant deviations from the SM (Collaboration, 2012, Collaboration, 2013). Model-dependent interpretations rule out vast swathes of parameter space in candidate theories:

• CMSSM/SUSY: Exclusion of gluino masses up to $\sim$710 GeV (for $m_0 > 1.3$ TeV in (Collaboration, 2012)), reaching above 1 TeV in later analyses and multilepton searches (Collaboration, 2013, Collaboration, 2020).
• Same-sign top production: Limits on $\sigma(pp \to tt + X)$ below $0.2\,\textrm{pb}$ depending on model (Lei, 2013, Collaboration, 2012).
• Four-top production: $\sigma_{4t}$ constrained below 49–85 fb (Lei, 2013, Collaboration, 2013).
• Vector-like quarks and fourth-generation scenarios: Mass exclusions extend up to the $0.6$–$0.9$ TeV range.
• RPV/MFV SUSY: Gluino limits exceeding $800$ GeV for generic stop masses via SSDL+b-jet searches (Berger et al., 2013).

The SSDL+jets signature is also a principal background in other new-physics searches, notably for Majorana behavior (seesaw neutrinos), doubly-charged Higgs, and leptoquarks. It is instrumental in developing charge-asymmetry observables, with recent studies exploring sensitivity to new effective operators and scalar sectors via $A_\mathrm{ch}$ differentials (Arganda et al., 2023).

7. Outlook and Future Directions

With increasing energy and luminosity at the LHC, the sensitivity of SSDL+jets analyses continues to rise. Projections for the HL-LHC indicate potential $5\sigma$ discovery or exclusion up to $2$–$3$ TeV for a variety of models, including singlet vectorlike top partners, 3-3-1 fermionic leptoquarks, and dimension-6 operators (Cui et al., 2022, Alves et al., 2012, Arganda et al., 2023). Methodological developments—such as advanced ML for fake-lepton discrimination, detailed charge-asymmetry studies, and sophisticated jet substructure for boosted-object separation—will further enhance coverage.

Continued data-driven constraints on SSDL+jets, improved background modeling, and enhanced recasting tools leveraging published efficiency parameterizations and kinematic templates (Collaboration, 2012, Collaboration, 2013, Collaboration, 2012) will be critical for discovery or exclusion in TeV-scale new physics searches. The SSDL+jets channel remains one of the most robust, low-background, and theoretically clean signatures in the quest to uncover physics beyond the Standard Model.