Supersymmetry searches in CMS Run 2: A complete review

TL;DR

The CMS Run 2 SUSY review surveys a colossal, signature-driven program that exploited $~$140 fb$^{-1}$ of data at $\sqrt{s}=13$ TeV to probe a broad range of SUSY scenarios. By combining inclusive and targeted analyses across gluino/squark, electroweakino, slepton, RPV, stealth, and LLP signatures, and by employing advanced object tagging, novel kinematic variables, and ML-based discrimination, CMS significantly tightened constraints on the SUSY landscape via both simplified-model and pMSSM interpretations. Although no excess was observed, the coordinated results yield a coherent picture of which SUSY regions are excluded and which remain viable, with the pMSSM study highlighting how current and external constraints sculpt the allowed parameter space and guiding future searches. The review also emphasizes methodological advances, data/analysis preservation, and the need for continued exploration in Run 3 and the HL-LHC era, where enhanced datasets and timing capabilities will extend sensitivity to challenging corners such as light higgsinos and compressed spectra.

Abstract

The Run 2 data-taking period of the CERN Large Hadron Collider during years 2015-2018 provided about 140 fb$^{-1}$ of proton-proton collisions at 13 TeV, offering an unprecedented opportunity to explore supersymmetry (SUSY) across a wide range of experimental signatures. CMS responded with a broad and diverse search program, carrying out dozens of analyses that probed a multitude of final states and systematically explored different regions of the SUSY parameter space. No significant deviations from standard model predictions were observed, and the results were used for constraining the SUSY landscape. In this review, I provide a comprehensive account of the CMS Run 2 SUSY program, covering its strategy, targeted models, and analysis methods. I then present the full set of searches and conclude with their combined impact through simplified model and phenomenological MSSM interpretations.

Figures (20)

• Figure 1: Diagrams for a nonexhaustive example set of simplified models covered by the inclusive searches.
• Figure 2: Highlights from inclusive analyses: Hadronic (left) and isolated and nonisolated leptonic (right) search regions and bins from the inclusive boosted analysis with razor variables.
• Figure 3: Diagrams for a nonexhaustive example set of simplified models covered by the top squark searches.
• Figure 4: Highlights from top squark searches: Complementarity of boosted and resolved top tagging efficiencies at function of generator level top quark $p_T$ from the fully hadronic top squark search (left); $p_T^{\text{miss}}\xspace$ distribution after preselection in the single lepton top squark search (center); and $M_{\text{T2}}(b\ell b\ell)$ distribution for the preselection from the dilepton top squark search (right).
• Figure 5: Diagrams for a nonexhaustive example set of simplified models covered by the electroweakino and slepton searches.