Pith. sign in

REVIEW 3 major objections 7 minor 124 references

The paper claims that machine learning on soft radiative photons can give the LHC a 5σ discovery reach for compressed singlino-higgsino dark matter up to a higgsino mass of 225 GeV, in a region where direct-detection experiments are blind.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review

2026-08-04 16:14 UTC pith:Y535BZ7I

load-bearing objection Solid and internally consistent ML projection of a motivated compressed NMSSM photon signature, but the quoted reach is statistics-only fast simulation and should be read as a search motivation, not a validated prediction. the 3 major comments →

arxiv 2509.15121 v2 pith:Y535BZ7I submitted 2025-09-18 hep-ph cs.LGhep-ex

Shedding Light on Dark Matter at the LHC with Machine Learning

Ernesto Arganda , Mart\'in de los Rios , Andres D. Perez , Subhojit Roy , Rosa M. Sand\'a Seoane , Carlos E. M. Wagner This is my paper
classification hep-ph cs.LGhep-ex PACS 95.35.+d12.60.Jv14.80.Ly
keywords dark matterNMSSMsinglino LSPelectroweakinoradiative neutralino decaycompressed mass spectrummachine learningLHC searches
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The authors investigate a dark matter candidate in the Z3-symmetric NMSSM: a singlino-dominated lightest neutralino that co-annihilates with nearby higgsino-like neutralinos. Because the mass splittings are only a few GeV, conventional three-body decays are suppressed and loop-induced radiative decays into photons dominate, yielding a distinctive but soft signature of one or two photons plus a lepton plus missing energy. They show that a machine-learned analysis of this final state, boosted by an initial-state radiation jet, can give a 5σ discovery for higgsino masses up to about 225 GeV with mass splittings below 12 GeV, using only 100 fb^-1 at 14 TeV. This reach covers parameter space that is inaccessible to current and future direct-detection experiments, including parts below the neutrino floor, so the LHC would be the only practical probe. The unbinned machine-learned likelihood method extends the binned-likelihood reach by about 10%.

Core claim

The paper's central claim is that radiative decays of higgsino-like neutralinos into a singlino-dominated LSP—χ~0_2 → χ~0_1 γ and χ~0_3 → χ~0_2 γ—can serve as a discovery channel for NMSSM dark matter at the LHC, even though the emitted photons and leptons are soft. In the compressed regime where the mass gap between the LSP and its co-annihilation partners is a few GeV, loop-induced radiative modes dominate over three-body decays, and the authors find that a gradient-boosted decision tree on the full event kinematics recovers the signal. With 100 fb^-1 at 14 TeV, they project a 5σ discovery for mχ~0_2 up to 225 GeV with mχ~0_2 − mχ~0_1 ≲ 12 GeV, and a 2σ exclusion up to 285 GeV with splitti

What carries the argument

The load-bearing mechanism is the compressed singlino–higgsino spectrum: for mass splitting ε, the conventional three-body neutralino decay is suppressed by ε^5 while the loop-induced radiative decay χ→χγ is suppressed only by ε^3, so photons dominate. Production proceeds through pp → χ~±_1 χ~0_2 j and pp → χ~±_1 χ~0_3 j, where the hard initial-state-radiation jet boosts the otherwise-soft decay products. A gradient-boosted decision tree is trained on seventeen kinematic variables (missing-energy significance, leading lepton and photon transverse momenta and transverse masses, object multiplicities, and related high-level features), and its continuous output feeds two statistical frameworks:

Load-bearing premise

The projected reach assumes the fast detector simulation faithfully reproduces real LHC acceptance for photons and leptons with transverse momentum down to 10 GeV, and that the listed Standard Model backgrounds are complete, with no unmodeled systematic uncertainties.

What would settle it

A dedicated 100 fb^-1 search at 14 TeV using the paper's event selection (at least one lepton, at least one photon, leading jet pT > 100 GeV, MET > 100 GeV) and its full machine-learning classifier would falsify the central claim if no excess appears where the paper predicts 5σ—most sharply at mχ~0_2 ≈ 225 GeV and Δm ≈ 12 GeV—provided the background estimate and detector acceptance match the simulation.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • A discovery (5σ) is projected for mχ~0_2 ≲ 225 GeV when the mass splitting to the LSP is ≲ 12 GeV, using only 100 fb^-1 at 14 TeV.
  • A 2σ exclusion extends to mχ~0_2 ≲ 285 GeV for splittings ≲ 20 GeV, covering regions beyond existing LHC multilepton bounds.
  • The pp→χ~±_1 χ~0_3 channel, with its sequential radiative decay χ~0_3→χ~0_2 γ→χ~0_1 γγ, is central: photon multiplicity is the third most-important classifier feature, and it keeps sensitivity alive even when the χ~0_2→χ~0_1 γ photon alone is too soft to pass the 10 GeV threshold.
  • The same parameter space is at or below the neutrino floor for direct detection, so this collider signature would be the only foreseeable way to probe it.
  • Binned and unbinned likelihood analyses agree, with the unbinned machine-learned likelihood extending the mass reach by about 10%.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Inference: Lowering the photon pT threshold below 10 GeV could extend the discovery region to mass splittings below 7 GeV, where the χ~0_2→χ~0_1 γ photon becomes very soft; the paper identifies this regime but does not quantify it.
  • Inference: Applied to already-recorded LHC data, the same classifier could look for a photon-enriched counterpart of the mild dilepton excesses seen in compressed electroweakino searches; the paper encourages such an analysis but does not perform it.
  • Inference: The unbinned machine-learned likelihood framework is model-agnostic; the same combination of soft-object kinematics and kernel-density-based likelihood would transfer to other compressed spectra, such as bino-wino or slepton co-annihilation scenarios.
  • Inference: The projected contours include only statistical uncertainties; a dedicated data-driven background estimate and systematic evaluation could shift the reach, and would be needed before using these numbers to plan an LHC search.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

3 major / 7 minor

Summary. The paper studies a compressed electroweakino spectrum in the Z3-symmetric NMSSM with a singlino-dominated LSP. In the direct-detection blind-spot region, the next-to-lightest states are higgsino-like and have enhanced radiative decays chi02 -> chi01 gamma and chi03 -> chi02 gamma; together with chargino production and an ISR jet, the final state is one soft lepton, one or two soft photons, and missing transverse energy. Signal and background events are generated with MadGraph5_aMC@NLO, Pythia8, and Delphes with the default ATLAS card. An XGBoost classifier is combined with a binned-likelihood and the unbinned machine-learned-likelihood (MLL) method. Using 100 fb^-1 at 14 TeV, the paper projects a 5-sigma discovery reach for m_chi02 up to about 225 GeV with Delta m <= 12 GeV and a 2-sigma exclusion reach up to about 285 GeV with Delta m <= 20 GeV, covering parameter space below the neutrino floor.

Significance. If the projection holds, the paper identifies a genuinely under-explored LHC channel for a well-motivated DM scenario that is difficult or impossible to probe with direct detection. The analysis has concrete strengths: fresh Monte Carlo samples for many benchmark points, explicit likelihood formulas (Eqs. 9-14), detailed tables of masses, branching ratios, cross sections, relic densities, and significances, and a recasting of existing LHC constraints with CheckMATE. These features make the main quantitative results reproducible in principle. The principal weakness is that the quoted reach is a fast-simulation, statistics-only projection; its robustness depends on unsupported assumptions about low-pT photon fakes, trigger performance, and background systematic uncertainties.

major comments (3)
  1. [Section 3, after the background list; Table 3] The background estimate includes W+jets, W gamma, ttbar+jets, etc., but contains no quantitative estimate of jet-to-photon fakes. The sentence 'Other sources were found to be negligible' is not supported by a control-region estimate or a fake-rate model. This is load-bearing because the signal photons are soft: for the benchmark splittings the rest-frame photon energy is about 6-8 GeV, and Figure 3 shows the leading-photon pT distribution peaking at low values. A modest fake rate could change the background count in the ML output and move the Z=2 and Z=5 contours in Figure 5 and hence the abstract's mass limits. Please quantify fakes, e.g., by applying an ATLAS-like jet-to-photon fake rate to simulated jets or by comparing with an alternative detector simulation, and state how the reach shifts; otherwise label the quoted reach as idealized.
  2. [Section 3.2, Eqs. (9)-(14); Figure 5] The quoted significances are purely statistical. The likelihoods contain only Poisson terms in the binned case and a Poisson factor plus KDE density ratios in the MLL case; the pseudo-experiments vary only Poisson counts. No nuisance parameters are included for background normalization, integrated luminosity, or selection efficiencies. For a 5-sigma claim this matters: a 20% background normalization uncertainty can reduce the effective significance by an amount comparable to the width of the stated contours, and the effect can be larger when many ML-output bins are used. Please profile over at least one background normalization nuisance parameter and show how the 225 GeV and 285 GeV contours change, or add a prominent caveat that the reach is an upper limit under zero systematics.
  3. [Section 3, Tables 3-4; Section 3.2] No trigger modeling is presented. The analysis requires at least one lepton with pT > 10 GeV and at least one photon with pT > 10 GeV, both below typical LHC single-lepton and diphoton trigger thresholds. The only hard objects are the leading jet with pT > 100 GeV and MET > 100 GeV. The reference in Section 3 to [152] addresses only the MET threshold and not the lepton/photon trigger legs. If the selection is presented as a realistic LHC search, the authors should demonstrate that the selected events pass a plausible trigger menu, e.g., a MET-plus-ISR-jet trigger with realistic turn-on efficiency, and quantify the resulting signal efficiency.
minor comments (7)
  1. [Abstract; Section 3] The paper quotes sqrt(s) = 14 TeV throughout, but LHC Run 3 operates at 13.6 TeV. Please clarify whether this is the design energy or an assumed future energy.
  2. [Section 2.4] The late-time entropy-injection assumption used to render overabundant benchmarks viable is stated in the text, but the abstract and conclusions do not carry this caveat. Please make visible in the summary that the DM interpretation of such benchmarks relies on an extra dilution mechanism beyond the NMSSM.
  3. [Section 3.2] The statement that for Delta m = m_chi02 - m_chi01 below about 7 GeV 'the significance starts to decrease again' is not consistently supported by Tables 7 and 8: some of the smallest-splitting points (e.g., BP1-4 and BP2-5) are the most significant in their series. Please rephrase or clarify that this trend is not uniform.
  4. [Tables 7 and 8] The chargino branching ratio is labelled BR(chi1pm -> chi01 Wpm), but the text says the W is off-shell. Use Wpm* consistently in the table headers or footnote.
  5. [Section 3] Generation details for the backgrounds (jet matching scheme, photon isolation in Delphes, and event weights) are not described. A short paragraph on these settings would aid reproducibility.
  6. [Figure 5] The gray region labeled 'LHC' is not described in the caption; please state that it comes from the CheckMATE/SModelS recasting and cite the relevant analyses in the caption.
  7. [Section 2.3] Typo: 'which corresponds to the the blind spot condition' should read 'the blind spot condition'.

Circularity Check

0 steps flagged

No circularity found: the projected reach is computed from fresh Monte Carlo plus a standard ML likelihood method; self-citations are methodological and non-load-bearing.

full rationale

The derivation chain from model to LHC reach is self-contained. The NMSSM scenario and blind-spot condition are taken from Ref. [42], an overlapping-author paper, but the current work explicitly states the blind-spot condition (Eq. 7) and independently computes the direct-detection cross sections with MicrOMEGAs for every benchmark point (Tables 7-8), so the DM-phenomenology premise is not assumed by citation alone. The ML tools (XGBoost, Binned Likelihood, Machine-Learned Likelihood) are cited to Refs. [51-53,152], but the reach contours in Fig. 5 are obtained from new MadGraph/Pythia/Delphessamples, with signal and background events generated separately and an independent test set used for evaluation (Section 3.1). No fitted parameter is renamed as a prediction: the significances Z_BL and Z_MLL are Monte Carlo projections, not fits to observed data. The self-citations that touch the analysis, e.g. Ref. [152] for the MET-threshold comment, are minor and do not determine the central mass-reach result by construction. The limitation noted in the conclusion ('a more in-depth and dedicated analysis of uncertainties and background modeling ... is worth pursuing') concerns systematic robustness and background fakes, not circularity. 'Other sources were found to be negligible' is an assumption, not a circular reduction. Thus no circular step is exhibited.

Axiom & Free-Parameter Ledger

4 free parameters · 6 axioms · 0 invented entities

No new particles, forces, or conserved quantities are introduced; the singlino, higgsinos, and singlet scalars are standard NMSSM content from the cited literature. The model inputs are chosen by hand (benchmark scanning) rather than fitted to data, and the paper is transparent about the dilution assumption. The main externally sourced pillars are the NMSSM framework itself and the radiative-decay dominance scaling, both from prior work.

free parameters (4)
  • M1 (bino soft mass) = 500 GeV
    Chosen to sit near the blind-spot condition (Eq. 7) and to enhance radiative transition rates. The paper states the collider reach is not significantly altered by this choice, but the branching ratios and direct-detection predictions depend on it.
  • kappa (singlet coupling) = 0.010-0.0133, varied per benchmark
    Scanned below 2*lambda to set the mass splitting between the singlino LSP and the higgsino-like states, thereby controlling the compression parameter epsilon and the radiative branching ratios.
  • mu_eff = 130-320 GeV, scanned
    Scanned to cover the higgsino mass range. Sets m_chi2_0 and, together with the jet requirement, determines the production cross sections quoted in Tables 7 and 8.
  • k-factor for electroweakino cross sections = 1.25
    Applied to all electroweakino production cross sections to approximate higher-order QCD corrections (Ref. [142]). A chosen normalization that affects the signal yield and therefore every quoted significance.
axioms (6)
  • domain assumption The Z3-symmetric NMSSM with superpotential W = W_MSSM|mu=0 + lambda S Hu Hd + (kappa/3) S^3 (Eq. 1) is the correct effective framework.
    The entire study is carried out inside this model; no independent evidence is offered that this is the actual beyond-Standard-Model sector.
  • domain assumption R-parity is conserved, making the lightest neutralino a stable dark-matter candidate.
    Section 2.1. Stability of the LSP is assumed, not derived, and the DM interpretation depends on it.
  • domain assumption For compressed spectra, three-body neutralino decays are suppressed as epsilon^5 while radiative decays scale as epsilon^3, so radiative modes dominate.
    Section 2.3, citing Ref. [114]. The dominance of chi_2^0 -> chi_1^0 gamma and chi_3^0 -> chi_2^0 gamma is the basis of the entire signature and is taken from prior literature rather than re-derived here.
  • ad hoc to paper Late-time entropy injection can dilute an overabundant neutralino relic density to the observed value without affecting collider observables.
    Section 2.4 states: 'We do not impose the Planck value as an equality constraint... allow late-time entropy injection... This assumption does not affect collider signatures.' This invocation keeps benchmark points with Omega h^2 > 0.12 viable.
  • domain assumption Delphes 3 with the default ATLAS configuration approximates the ATLAS detector for soft photons, leptons, and jets.
    Section 3, event simulation. All signal acceptances, background rates, and resulting significances depend on this approximation, including the low-pT photon response.
  • standard math The Asimov and pseudo-experiment machinery gives valid test statistics for the binned and machine-learned likelihoods.
    Section 3.1, Eqs. (9)-(14). This is standard statistics, but the MLL-specific KDE densities and the choice of mu'=1 pseudo-experiments are validated only by reference to the authors' method papers.

pith-pipeline@v1.3.0-alltime-deepseek · 31377 in / 18975 out tokens · 207434 ms · 2026-08-04T16:14:17.121987+00:00 · methodology

0 comments
read the original abstract

We investigate a WIMP dark matter (DM) candidate in the form of a singlino-dominated lightest supersymmetric particle (LSP) within the $Z_3$-symmetric Next-to-Minimal Supersymmetric Standard Model (NMSSM). This framework gives rise to regions of parameter space where DM is obtained via co-annihilation with nearby higgsino-like electroweakinos and DM direct detection~signals are suppressed, the so-called ``blind spots''. On the other hand, collider signatures remain promising due to enhanced radiative decay modes of higgsinos into the singlino-dominated LSP and photons, rather than into leptons or hadrons. Compared to MSSM scenarios with light bino- and wino-like electroweakinos, the NMSSM allows for final states with multiple photons arising from cascade radiative decays, providing a distinctive collider signature. This motivates searches for radiatively decaying neutralinos, however, these signals face substantial background challenges, as the decay products are typically soft due to the small mass-splits ($\Delta m$) between the LSP and the higgsino-like coannihilation partners. We apply a data-driven Machine Learning (ML) analysis that improves sensitivity to these subtle signals, offering a powerful complement to traditional search strategies to discover a new physics scenario. Using an LHC integrated luminosity of $100~\mathrm{fb}^{-1}$ at $14~\mathrm{TeV}$, the method achieves a $5\sigma$ discovery reach for higgsino masses up to $225~\mathrm{GeV}$ with $\Delta m\!\lesssim\!12~\mathrm{GeV}$, and a $2\sigma$ exclusion up to $285~\mathrm{GeV}$ with $\Delta m\!\lesssim\!20~\mathrm{GeV}$. These results highlight~the power of collider searches to probe DM candidates that remain hidden from current~direct detection experiments, and provide a motivation for a search by the LHC collaborations using ML methods.

Figures

Figures reproduced from arXiv: 2509.15121 by Andres D. Perez, Carlos E. M. Wagner, Ernesto Arganda, Mart\'in de los Rios, Rosa M. Sand\'a Seoane, Subhojit Roy.

Figure 1
Figure 1. Figure 1: Feynman diagrams of the two production channel modes of the higgsino-like states, [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Left panel: ROC curve and AUC performance metric of the XGBoost classifier trained using all [PITH_FULL_IMAGE:figures/full_fig_p015_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Distributions of the six final state kinematic variables showing the most discrimination between [PITH_FULL_IMAGE:figures/full_fig_p016_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: 2D distributions between some of the most important features in the classification task. Green [PITH_FULL_IMAGE:figures/full_fig_p018_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Projected discovery significance Z = 2 (dashed) and Z = 5 (solid) contour lines in the [mχ˜ 0 2 , mχ˜ 0 2 − mχ˜ 0 1 ] plane applying the Binned-Likelihood approach (BL, orange), and the Machine-Learned Likelihood method (MLL, violet). lower mass splittings mχ˜ 0 2 − mχ˜ 0 1 ). Note that even for electroweakinos with mass splittings below 10 GeV (the lower limit of the photon identification criteria) we hav… view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

124 extracted references · 106 linked inside Pith

  1. [3]

    Fayet,Supersymmetry and Weak, Electromagnetic and Strong Interactions,Phys

    P. Fayet,Supersymmetry and Weak, Electromagnetic and Strong Interactions,Phys. Lett. B 64(1976) 159

    1976

  2. [4]

    Fayet,Spontaneously Broken Supersymmetric Theories of Weak, Electromagnetic and Strong Interactions,Phys

    P. Fayet,Spontaneously Broken Supersymmetric Theories of Weak, Electromagnetic and Strong Interactions,Phys. Lett. B69(1977) 489

    1977

  3. [5]

    Nilles,Supersymmetry, Supergravity and Particle Physics,Phys

    H.P. Nilles,Supersymmetry, Supergravity and Particle Physics,Phys. Rept.110(1984) 1

    1984

  4. [6]

    Haber and G.L

    H.E. Haber and G.L. Kane,The Search for Supersymmetry: Probing Physics Beyond the Standard Model,Phys. Rept.117(1985) 75

    1985

  5. [7]

    Gunion and H.E

    J. Gunion and H.E. Haber,Higgs Bosons in Supersymmetric Models. 1.,Nucl. Phys. B272 (1986) 1

    1986

  6. [8]

    Martin,A Supersymmetry primer,Adv

    S.P. Martin,A Supersymmetry primer,Adv. Ser. Direct. High Energy Phys.18(1998) 1 [hep-ph/9709356]

    Pith/arXiv arXiv 1998

  7. [9]

    Farrar and P

    G.R. Farrar and P. Fayet,Phenomenology of the Production, Decay, and Detection of New Hadronic States Associated with Supersymmetry,Phys. Lett. B76(1978) 575

    1978

  8. [10]

    Dimopoulos and H

    S. Dimopoulos and H. Georgi,Softly Broken Supersymmetry and SU(5),Nucl. Phys. B193 (1981) 150

    1981

  9. [11]

    Weinberg,Supersymmetry at Ordinary Energies

    S. Weinberg,Supersymmetry at Ordinary Energies. 1. Masses and Conservation Laws,Phys. Rev. D26(1982) 287

    1982

  10. [12]

    Sakai and T

    N. Sakai and T. Yanagida,Proton Decay in a Class of Supersymmetric Grand Unified Models,Nucl. Phys. B197(1982) 533

    1982

  11. [13]

    Dimopoulos, S

    S. Dimopoulos, S. Raby and F. Wilczek,Proton Decay in Supersymmetric Models,Phys. Lett. B112(1982) 133

    1982

  12. [14]

    Larkoski, I

    A.J. Larkoski, I. Moult and B. Nachman,Jet Substructure at the Large Hadron Collider: A Review of Recent Advances in Theory and Machine Learning,Phys. Rept.841(2020) 1 [1709.04464]

    Pith/arXiv arXiv 2020

  13. [15]

    Mehta, M

    P. Mehta, M. Bukov, C.-H. Wang, A.G.R. Day, C. Richardson, C.K. Fisher et al.,A high-bias, low-variance introduction to Machine Learning for physicists,Phys. Rept.810 (2019) 1 [1803.08823]

    Pith/arXiv arXiv 2019

  14. [16]

    Guest, K

    D. Guest, K. Cranmer and D. Whiteson,Deep Learning and its Application to LHC Physics, Ann. Rev. Nucl. Part. Sci.68(2018) 161 [1806.11484]

    Pith/arXiv arXiv 2018

  15. [17]

    Albertsson et al.,Machine Learning in High Energy Physics Community White Paper,J

    K. Albertsson et al.,Machine Learning in High Energy Physics Community White Paper,J. Phys. Conf. Ser.1085(2018) 022008 [1807.02876]

    Pith/arXiv arXiv 2018

  16. [18]

    Radovic, M

    A. Radovic, M. Williams, D. Rousseau, M. Kagan, D. Bonacorsi, A. Himmel et al.,Machine learning at the energy and intensity frontiers of particle physics,Nature560(2018) 41. 25

    2018

  17. [19]

    Carleo, I

    G. Carleo, I. Cirac, K. Cranmer, L. Daudet, M. Schuld, N. Tishby et al.,Machine learning and the physical sciences,Rev. Mod. Phys.91(2019) 045002 [1903.10563]

    Pith/arXiv arXiv 2019

  18. [20]

    Bourilkov,Machine and Deep Learning Applications in Particle Physics,Int

    D. Bourilkov,Machine and Deep Learning Applications in Particle Physics,Int. J. Mod. Phys. A34(2020) 1930019 [1912.08245]

    Pith/arXiv arXiv 2020

  19. [21]

    Feickert and B

    M. Feickert and B. Nachman,A Living Review of Machine Learning for Particle Physics, 2102.02770

    Pith/arXiv arXiv

  20. [22]

    Schwartz,Modern Machine Learning and Particle Physics,2103.12226

    M.D. Schwartz,Modern Machine Learning and Particle Physics,2103.12226

    Pith/arXiv arXiv

  21. [23]

    Karagiorgi, G

    G. Karagiorgi, G. Kasieczka, S. Kravitz, B. Nachman and D. Shih,Machine Learning in the Search for New Fundamental Physics,2112.03769

    Pith/arXiv arXiv

  22. [24]

    Coadou,Boosted decision trees,2206.09645

    Y. Coadou,Boosted decision trees,2206.09645

    Pith/arXiv arXiv

  23. [25]

    Shanahan et al.,Snowmass 2021 Computational Frontier CompF03 Topical Group Report: Machine Learning,2209.07559

    P. Shanahan et al.,Snowmass 2021 Computational Frontier CompF03 Topical Group Report: Machine Learning,2209.07559

    Pith/arXiv arXiv 2021

  24. [26]

    Plehn, A

    T. Plehn, A. Butter, B. Dillon and C. Krause,Modern Machine Learning for LHC Physicists,2211.01421

    Pith/arXiv arXiv

  25. [27]

    Belis, P

    V. Belis, P. Odagiu and T.K. Aarrestad,Machine learning for anomaly detection in particle physics,Rev. Phys.12(2024) 100091 [2312.14190]

    Pith/arXiv arXiv 2024

  26. [28]

    Bardhan, T

    J. Bardhan, T. Mandal, S. Mitra, C. Neeraj and M. Patra,Unsupervised and lightly supervised learning in particle physics,2403.13676

    Pith/arXiv arXiv

  27. [29]

    Mondal and L

    S. Mondal and L. Mastrolorenzo,Machine Learning in High Energy Physics: A review of heavy-flavor jet tagging at the LHC,2404.01071

    Pith/arXiv arXiv

  28. [30]

    Halverson,TASI Lectures on Physics for Machine Learning,2408.00082

    J. Halverson,TASI Lectures on Physics for Machine Learning,2408.00082

    Pith/arXiv arXiv

  29. [31]

    Ellwanger, C

    U. Ellwanger, C. Hugonie and A.M. Teixeira,The Next-to-Minimal Supersymmetric Standard Model,Phys. Rept.496(2010) 1 [0910.1785]

    Pith/arXiv arXiv 2010

  30. [32]

    Steigman and M.S

    G. Steigman and M.S. Turner,Cosmological Constraints on the Properties of Weakly Interacting Massive Particles,Nucl. Phys. B253(1985) 375

    1985

  31. [33]

    Arcadi, M

    G. Arcadi, M. Dutra, P. Ghosh, M. Lindner, Y. Mambrini, M. Pierre et al.,The waning of the WIMP? A review of models, searches, and constraints,Eur. Phys. J. C78(2018) 203 [1703.07364]. [34]Planckcollaboration,Planck 2018 results. VI. Cosmological parameters,Astron. Astrophys.641(2020) A6 [1807.06209]

    Pith/arXiv arXiv 2018

  32. [35]

    Arkani-Hamed, A

    N. Arkani-Hamed, A. Delgado and G.F. Giudice,The Well-tempered neutralino,Nucl. Phys. B741(2006) 108 [hep-ph/0601041]. [36]XENONcollaboration,First Dark Matter Search with Nuclear Recoils from the XENONnT Experiment,Phys. Rev. Lett.131(2023) 041003 [2303.14729]. [37]PandaX-4Tcollaboration,Dark Matter Search Results from the PandaX-4T Commissioning Run,Phy...

    Pith/arXiv arXiv 2006

  33. [40]

    Badziak, M

    M. Badziak, M. Olechowski and P. Szczerbiak,Blind spots for neutralino dark matter in the NMSSM,JHEP03(2016) 179 [1512.02472]

    Pith/arXiv arXiv 2016

  34. [41]

    S. Baum, M. Carena, N.R. Shah and C.E.M. Wagner,Higgs portals for thermal Dark Matter. EFT perspectives and the NMSSM,JHEP04(2018) 069 [1712.09873]

    Pith/arXiv arXiv 2018

  35. [42]

    Roy and C.E.M

    S. Roy and C.E.M. Wagner,Dark Matter searches with photons at the LHC,JHEP04 (2024) 106 [2401.08917]. [43]ATLAScollaboration,Search for chargino-neutralino pair production in final states with three leptons and missing transverse momentum in√s= 13 TeVppcollisions with the ATLAS detector,ATLAS-CONF-2020-015(2020) . [44]CMScollaboration,Search for electrowe...

    Pith/arXiv arXiv 2024

  36. [50]

    Cowan, K

    G. Cowan, K. Cranmer, E. Gross and O. Vitells,Asymptotic formulae for likelihood-based tests of new physics,Eur. Phys. J. C71(2011) 1554 [1007.1727]

    Pith/arXiv arXiv 2011

  37. [51]

    Arganda, X

    E. Arganda, X. Marcano, V.M. Lozano, A.D. Medina, A.D. Perez, M. Szewc et al.,A method for approximating optimal statistical significances with machine-learned likelihoods, Eur. Phys. J. C82(2022) 993 [2205.05952]

    Pith/arXiv arXiv 2022

  38. [52]

    Arganda, M

    E. Arganda, M. de los Rios, A.D. Perez and R.M. Sandá Seoane,Imposing exclusion limits on new physics with machine-learned likelihoods,PoSICHEP2022(2022) 1226

    2022

  39. [53]

    Arganda, A.D

    E. Arganda, A.D. Perez, M. de los Rios and R.M. Sandá Seoane,Machine-learned exclusion limits without binning,Eur. Phys. J. C83(2023) 1158 [2211.04806]

    Pith/arXiv arXiv 2023

  40. [54]

    Arganda, D.A

    E. Arganda, D.A. Díaz, A.D. Perez, R.M. Sandá Seoane and A. Szynkman,LHC study of third-generation scalar leptoquarks with machine-learned likelihoods,Phys. Rev. D109 (2024) 055032 [2309.05407]

    Pith/arXiv arXiv 2024

  41. [55]

    Rosenblatt,Remarks on Some Nonparametric Estimates of a Density Function,Ann

    M. Rosenblatt,Remarks on Some Nonparametric Estimates of a Density Function,Ann. Math. Statist.27(1956) 832. 27

    1956

  42. [56]

    Parzen,On Estimation of a Probability Density Function and Mode,Ann

    E. Parzen,On Estimation of a Probability Density Function and Mode,Ann. Math. Statist. 33(1962) 1065

    1962

  43. [57]

    J. Cao, Y. He, L. Shang, W. Su and Y. Zhang,Natural NMSSM after LHC Run I and the Higgsino dominated dark matter scenario,JHEP08(2016) 037 [1606.04416]

    Pith/arXiv arXiv 2016

  44. [58]

    Ellwanger,Present Status and Future Tests of the Higgsino-Singlino Sector in the NMSSM,JHEP02(2017) 051 [1612.06574]

    U. Ellwanger,Present Status and Future Tests of the Higgsino-Singlino Sector in the NMSSM,JHEP02(2017) 051 [1612.06574]

    Pith/arXiv arXiv 2017

  45. [59]

    Xiang, X.-J

    Q.-F. Xiang, X.-J. Bi, P.-F. Yin and Z.-H. Yu,Searching for Singlino-Higgsino Dark Matter in the NMSSM,Phys. Rev. D94(2016) 055031 [1606.02149]

    Pith/arXiv arXiv 2016

  46. [60]

    J. Cao, Y. He, L. Shang, Y. Zhang and P. Zhu,Current status of a natural NMSSM in light of LHC 13 TeV data and XENON-1T results,Phys. Rev. D99(2019) 075020 [1810.09143]

    Pith/arXiv arXiv 2019

  47. [61]

    Ellwanger and C

    U. Ellwanger and C. Hugonie,The higgsino–singlino sector of the NMSSM: combined constraints from dark matter and the LHC,Eur. Phys. J. C78(2018) 735 [1806.09478]

    Pith/arXiv arXiv 2018

  48. [62]

    Domingo, J.S

    F. Domingo, J.S. Kim, V.M. Lozano, P. Martin-Ramiro and R. Ruiz de Austri,Confronting the neutralino and chargino sector of the NMSSM with the multilepton searches at the LHC, Phys. Rev. D101(2020) 075010 [1812.05186]

    Pith/arXiv arXiv 2020

  49. [63]

    Baum, N.R

    S. Baum, N.R. Shah and K. Freese,The NMSSM is within Reach of the LHC: Mass Correlations \& Decay Signatures,JHEP04(2019) 011 [1901.02332]

    Pith/arXiv arXiv 2019

  50. [64]

    van Beekveld, S

    M. van Beekveld, S. Caron and R. Ruiz de Austri,The current status of fine-tuning in supersymmetry,JHEP01(2020) 147 [1906.10706]

    Pith/arXiv arXiv 2020

  51. [65]

    Abdallah, A

    W. Abdallah, A. Chatterjee and A. Datta,Revisiting singlino dark matter of the natural Z3-symmetric NMSSM in the light of LHC,JHEP09(2019) 095 [1907.06270]

    Pith/arXiv arXiv 2019

  52. [66]

    J. Cao, L. Meng, Y. Yue, H. Zhou and P. Zhu,Suppressing the scattering of WIMP dark matter and nucleons in supersymmetric theories,Phys. Rev. D101(2020) 075003 [1910.14317]

    Pith/arXiv arXiv 2020

  53. [67]

    Abdallah, A

    W. Abdallah, A. Datta and S. Roy,A relatively light, highly bino-like dark matter in the Z3-symmetric NMSSM and recent LHC searches,JHEP04(2021) 122 [2012.04026]

    Pith/arXiv arXiv 2021

  54. [68]

    Lopez-Fogliani, A.D

    D.E. Lopez-Fogliani, A.D. Perez and R.R. de Austri,Dark matter candidates in the NMSSM with RH neutrino superfields,JCAP04(2021) 067 [2102.08986]

    Pith/arXiv arXiv 2021

  55. [69]

    Kim, D.E

    J.S. Kim, D.E. Lopez-Fogliani, A.D. Perez and R.R. de Austri,The new (g−2)µand right-handed sneutrino dark matter,Nucl. Phys. B974(2022) 115637 [2107.02285]

    Pith/arXiv arXiv 2022

  56. [70]

    Kim, D.E

    J.S. Kim, D.E. Lopez-Fogliani, A.D. Perez and R.R. de Austri,Right-handed sneutrino and gravitino multicomponent dark matter in light of neutrino detectors,JCAP04(2023) 050 [2206.04715]

    Pith/arXiv arXiv 2023

  57. [71]

    Datta, M

    A. Datta, M. Guchait, A. Roy and S. Roy,Hunting ewinos and a light scalar of Z3-NMSSM with a bino-like dark matter in top squark decays at the LHC,JHEP11(2023) 081 [2211.05905]

    Pith/arXiv arXiv 2023

  58. [72]

    D. Das, U. Ellwanger and A.M. Teixeira,Modified Signals for Supersymmetry in the NMSSM with a Singlino-like LSP,JHEP04(2012) 067 [1202.5244]

    Pith/arXiv arXiv 2012

  59. [73]

    Ellwanger and A.M

    U. Ellwanger and A.M. Teixeira,NMSSM with a singlino LSP: possible challenges for searches for supersymmetry at the LHC,JHEP10(2014) 113 [1406.7221]. 28

    Pith/arXiv arXiv 2014

  60. [74]

    Chatterjee, A

    A. Chatterjee, A. Datta and S. Roy,Electroweak phase transition in the Z3-invariant NMSSM: Implications of LHC and Dark matter searches and prospects of detecting the gravitational waves,JHEP06(2022) 108 [2202.12476]

    Pith/arXiv arXiv 2022

  61. [75]

    J. Cao, F. Li, J. Lian, Y. Pan and D. Zhang,Impact of LHC probes of SUSY and recent measurement of (g−2) µ onZ 3-NMSSM,Sci. China Phys. Mech. Astron.65(2022) 291012 [2204.04710]

    Pith/arXiv arXiv 2022

  62. [76]

    J. Cao, L. Meng and Y. Yue,Electron and muon anomalous magnetic moments in the Z3-NMSSM,Phys. Rev. D108(2023) 035043 [2306.06854]

    Pith/arXiv arXiv 2023

  63. [77]

    Bisal and D

    S. Bisal and D. Das,Testing Z boson rare decaysZ→H1γ, A1γwith(g−2) µ, MW ,and BR(hSM →Zγ)in the NMSSM,Eur. Phys. J. C84(2024) 630 [2308.06558]

    Pith/arXiv arXiv 2024

  64. [78]

    Guchait and A

    M. Guchait and A. Roy,Light Singlino Dark Matter at the LHC,Phys. Rev. D102(2020) 075023 [2005.05190]

    Pith/arXiv arXiv 2020

  65. [79]

    J. Cao, X. Jia, L. Meng, Y. Yue and D. Zhang,Status of the singlino-dominated dark matter in general Next-to-Minimal Supersymmetric Standard Model,JHEP03(2023) 198 [2210.08769]

    Pith/arXiv arXiv 2023

  66. [80]

    Adhikary, R.K

    A. Adhikary, R.K. Barman, B. Bhattacherjee, A. De and R.M. Godbole,Exploring the Singlino-dominated Thermal Neutralino Dark Matter in theZ3 invariant NMSSM, 2408.03656

    Pith/arXiv arXiv

  67. [81]

    L. Meng, J. Cao, F. Li and S. Yang,Dark Matter physics in general NMSSM,JHEP08 (2024) 212 [2405.07036]

    Pith/arXiv arXiv 2024

  68. [82]

    Cheung, M

    C. Cheung, M. Papucci, D. Sanford, N.R. Shah and K.M. Zurek,NMSSM Interpretation of the Galactic Center Excess,Phys. Rev. D90(2014) 075011 [1406.6372]

    Pith/arXiv arXiv 2014

  69. [83]

    Badziak, M

    M. Badziak, M. Olechowski and P. Szczerbiak,Spin-dependent constraints on blind spots for thermal singlino-higgsino dark matter with(out) light singlets,JHEP07(2017) 050 [1705.00227]

    Pith/arXiv arXiv 2017

  70. [84]

    Ellwanger, J.F

    U. Ellwanger, J.F. Gunion and C. Hugonie,NMHDECAY: A Fortran code for the Higgs masses, couplings and decay widths in the NMSSM,JHEP02(2005) 066 [hep-ph/0406215]

    Pith/arXiv arXiv 2005

  71. [85]

    J. Fan, M. Reece and J.T. Ruderman,Stealth Supersymmetry,JHEP11(2011) 012 [1105.5135]

    Pith/arXiv arXiv 2011

  72. [86]

    Ellwanger, M

    U. Ellwanger, M. Rausch de Traubenberg and C.A. Savoy,Phenomenology of supersymmetric models with a singlet,Nucl. Phys. B492(1997) 21 [hep-ph/9611251]

    Pith/arXiv arXiv 1997

  73. [87]

    Belanger, F

    G. Belanger, F. Boudjema, C. Hugonie, A. Pukhov and A. Semenov,Relic density of dark matter in the NMSSM,JCAP09(2005) 001 [hep-ph/0505142]

    Pith/arXiv arXiv 2005

  74. [88]

    Gunion, D

    J.F. Gunion, D. Hooper and B. McElrath,Light neutralino dark matter in the NMSSM, Phys. Rev. D73(2006) 015011 [hep-ph/0509024]

    Pith/arXiv arXiv 2006

  75. [89]

    Cheung and D

    C. Cheung and D. Sanford,Simplified Models of Mixed Dark Matter,JCAP02(2014) 011 [1311.5896]

    Pith/arXiv arXiv 2014

  76. [90]

    J. Fan, M. Reece and J.T. Ruderman,A Stealth Supersymmetry Sampler,JHEP07(2012) 196 [1201.4875]. 29

    Pith/arXiv arXiv 2012

  77. [91]

    Berlin, P

    A. Berlin, P. Gratia, D. Hooper and S.D. McDermott,Hidden Sector Dark Matter Models for the Galactic Center Gamma-Ray Excess,Phys. Rev. D90(2014) 015032 [1405.5204]

    Pith/arXiv arXiv 2014

  78. [92]

    Draper, T

    P. Draper, T. Liu, C.E.M. Wagner, L.-T. Wang and H. Zhang,Dark Light Higgs,Phys. Rev. Lett.106(2011) 121805 [1009.3963]

    Pith/arXiv arXiv 2011

  79. [93]

    Cerdeno, C

    D.G. Cerdeno, C. Hugonie, D.E. Lopez-Fogliani, C. Munoz and A.M. Teixeira,Theoretical predictions for the direct detection of neutralino dark matter in the NMSSM,JHEP12 (2004) 048 [hep-ph/0408102]

    Pith/arXiv arXiv 2004

  80. [94]

    J. Cao, L. Shang, P. Wu, J.M. Yang and Y. Zhang,Interpreting the galactic center gamma-ray excess in the NMSSM,JHEP10(2015) 030 [1506.06471]

    Pith/arXiv arXiv 2015

Showing first 80 references.

This paper was first reviewed by deepseek-v4-flash on August 4, 2026.