Pith. sign in

REVIEW 4 major objections 5 minor 62 references

A convolutional neural network can infer the presence of two dark matter particles from LHC mono-jet and mono-Z events, then estimate their masses and, in one-component signals, distinguish a spin-0 scalar from a spin-1/2 fermion.

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 · deepseek-v4-flash

2026-08-05 04:47 UTC pith:4TH4KP3J

load-bearing objection A transparent, genuinely useful proof-of-concept for characterizing two-component dark matter with a CNN, but the mono-Z spin claim is undermined by disjoint fermion/scalar mass ranges in the training data and needs an overlapping-mass control. the 4 major comments →

arxiv 2608.03975 v1 pith:4TH4KP3J submitted 2026-08-04 hep-ph

Mono-X Signal Characterization from Two-component Dark Matter Using a Convolutional Neural Network

Max Fust\'e Costa , Yong Sheng Koay , Stefano Moretti This is my paper
classification hep-ph
keywords two-component dark mattermono-jetmono-Zconvolutional neural networkLHCdark matter mass regressiondark matter spinmissing transverse energy
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 paper tries to establish that a convolutional neural network (CNN) can do post-discovery characterization for a dark sector containing two particles, not just one. Using simulated mono-jet and mono-Z events from a minimal model with one fermion and one scalar dark matter candidate, it shows that a 1D CNN fed with binned kinematic distributions can first classify a signal as one- or two-component, and then regress the dark matter mass(es); with the mono-Z channel it can also tell the spin-0 scalar from the spin-1/2 fermion in a one-component signal. The point is that if the LHC ever sees dark matter, the next question is whether the dark sector has several particles and what their masses and spins are—exactly the information this network is trained to output. These results are explicitly a proof of concept: the analysis assumes a background-free signal and comparable production cross sections for the two components, so the quoted accuracies are best-case numbers.

Core claim

The central claim is that the same CNN architecture can solve three related characterization tasks on distributions reconstructed at detector level. On mono-jet data (transverse momentum, missing transverse energy, pseudorapidity) the classifier separates one-component from two-component dark matter signals with roughly 70% accuracy on the two-component class, and the regressors recover the masses: within about 50–80 GeV at 68% tolerance for single-component scalar/fermion signals and about 100 GeV for two-component signals. On mono-Z data, the same network also uses the angular separation Δφ between the two leptons to distinguish a single scalar from a single fermion, and mass estimates tig

What carries the argument

The engine of the analysis is an n-observable one-dimensional CNN (n-1DCNN). Each reconstructed kinematic variable—pT, missing transverse energy, η, and (for mono-Z) Δφ—is binned into a histogram, and these histograms are fed as separate channels into 1D convolutional layers with dropout and max-pooling, followed by flattening and dense layers so correlations between observables can be re-learned. The design is deliberately extensible to more observables or more final states. Around it, two choices make the learning problem well-posed: the training data are normalized shape-only distributions, and the model parameters are restricted (via a cross-section-ratio condition and coupling rescaling

Load-bearing premise

The results assume a signal-only environment in which both dark matter components are produced with nearly equal cross sections (within 15%) and the couplings are rescaled per mass point; if real data contains Standard Model backgrounds or one component dominates production, the network's classifications and mass estimates are not guaranteed to transfer.

What would settle it

Retrain or re-evaluate the same classifier and regressors on samples that include the usual LHC backgrounds (Z→νν+jets, W→lν+jets) or on samples with a 5:1 cross-section imbalance between the two components (the r=5 case the paper generated but did not report). If the two-component classification accuracy and the stated 68% mass tolerances fall to chance level under either change, the background-free, balanced-training condition is the actual driver of the claim.

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

Share X Bluesky LinkedIn Reddit HN

If this is right

  • If the claim holds, mono-jet data alone is enough to determine that a would-be dark matter signal contains two components and to give a rough mass for each (of order 100 GeV at 68% tolerance).
  • Mono-Z adds the spin axis: a one-component signal can be assigned to a spin-1/2 fermion or a spin-0 scalar, and mass precision improves, especially for the scalar.
  • Because the CNN consumes histograms, it is a natural fit for detector-level summary statistics and can be extended to more observables (e.g., additional jets or leptons) without architectural redesign.
  • The two-stage workflow (classify first, then regress) means the spin and component-count decisions are made once, and the mass estimators never have to hedge across classes.
  • The quoted precisions are best-case, signal-only values; a realistic analysis with backgrounds can reasonably expect worse performance, as the paper itself warns.

Where Pith is reading between the lines

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

  • A natural extension would be to inject standard-model backgrounds (Z→νν+jets, W→lν+jets, top-quark pairs) and evaluate how the 70% and 68% numbers degrade; the paper's own caveat implies they will.
  • The two-component mono-Z mass bias toward low values suggests kinematic degeneracy or partial dominance of one component; a test is to train on varied cross-section ratios r=2, 3, 5 (the paper generated such points but reports mainly r=1) and see whether the bias tracks r.
  • Spin discrimination is only demonstrated for one-component signals; in a mixed two-component sample, the Δφ distribution would contain contributions from both spin channels, and disentangling them is a further problem the present architecture does not address.
  • The histogram-based n-1DCNN also suggests a way to use the same data representation for simulation-based inference (likelihood-free posteriors) rather than point estimates, which the paper explicitly leaves for future work.

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

4 major / 5 minor

Summary. This paper studies a two-component DM model with a Dirac fermion chi and a real scalar S plus a scalar mediator. The authors generate LO mono-jet and mono-Z events with MadGraph/Pythia/Delphes, normalize histograms of pT, MET, eta (and Delta-phi for mono-Z), and train a multi-channel 1D CNN to (i) classify events as one-component fermion, one-component scalar, or two-component DM; (ii) regress the DM mass(es). All results are presented under a background-free hypothesis and the paper is explicitly framed as a proof-of-concept (Abstract; Sec. 1; Sec. 4; Sec. 5). The main reported findings are: mono-jet classifiers distinguish two-component from one-component signals with roughly 70% recall for the two-component class (Fig. 13); mono-Z classifiers also separate one-component fermion from scalar; mass regression achieves best performance for one-component mono-Z and degrades substantially for two-component signals, with 68% tolerances of order 100 GeV in mono-jet and 75-125 GeV in mono-Z, and 90% tolerances of 150-250 GeV. The authors conclude that a CNN can infer the presence, mass, and spin of DM components.

Significance. Demonstrating that a CNN can extract DM spin in addition to mass from mono-X events would be a valuable addition to the ML-based dark-sector characterization literature, complementing previous one-component studies (Refs. [17,18]). The paper has real strengths: a detector-level simulation chain, a clearly described network architecture, results averaged over 50 independent training/validation splits, and an explicit statement of the background-free proof-of-concept limitation. However, the central novelty--spin extraction from mono-Z data--is not established because the training samples for fermion and scalar DM have disjoint mass ranges (Table 1). The classification and regression numbers also need interpretation with care: the two-component mono-jet class has only ~70% recall and the two-component mass regressions have 90% tolerances of up to 250 GeV. Because the spin claim is load-bearing and currently confounded, the paper cannot be accepted in its present form.

major comments (4)
  1. [Table 1 / Sec. 3.1.2 / Sec. 4.2] The mono-Z spin-discrimination result is confounded by the training-data mass ranges. Table 1 lists the one-component mono-Z scalar samples with pole masses in [80,230] GeV and the fermion samples in [230,1000] GeV. These ranges are disjoint apart from a single boundary value. A classifier trained to label 'fermion' vs 'scalar' can therefore separate the classes using mass-correlated kinematic variables (lepton pT, MET, eta, Delta-phi) without learning any spin-dependent feature. The footnote to Table 1 notes that mass ranges differ between one- and two-component cases but does not acknowledge that the fermion and scalar one-component ranges are disjoint. Without a mass-matched test, e.g., overlapping mass windows for both spins, the abstract's claim of 'extracting their spin' is not supported.
  2. [Sec. 4.2 / Fig. 13] The mono-jet classifier shows roughly 70% recall for true two-component events, with ~30% misclassified as one-component (Fig. 13). The text does not report overall accuracy or per-class precision/recall, yet the conclusions describe the classification as 'high accuracy'. Since 'inferring the presence of two DM particles' is a central claim, the paper should report full quantitative classification metrics with uncertainties over the 50 runs and temper the wording accordingly.
  3. [Sec. 3.1.1 / Eqs. (8),(10) / Sec. 4.1 footnote] All presented results correspond to r=1, i.e., fermion and scalar cross sections within 15% of each other (Eq. (8)). The paper states that r=2,3,5 data sets were also generated (footnote in Sec. 4.1), but no results for those cases are shown. The practical utility for a generic two-component signal depends on robustness when one component dominates; at minimum the authors should explain why the r>1 data are omitted or add a sensitivity study. Relatedly, since all histograms are normalized to unit area (Sec. 3.2.1), absolute rate information that could help separate components in a real analysis is discarded; this should be stated as an explicit limitation of the proof-of-concept.
  4. [Sec. 4.3.2 / Figs. 22,25] Mass extraction in the two-component case is quantitatively modest: mono-jet 68% tolerance is about 100 GeV and 90% tolerance about 250 GeV; mono-Z 68% tolerances are 75 GeV (scalar) and 125 GeV (fermion), with 90% at roughly 225 GeV for the fermion. These numbers should be stated in the abstract or conclusions when claiming that the CNN can 'extract' the DM masses, so that readers can judge the precision. The point estimates are also presented without calibrated uncertainties, which makes the regression claim difficult to interpret.
minor comments (5)
  1. [Title / Sec. 4.1] Typos and formatting: 'Mono-XSignal' in the title should be 'Mono-X Signal'; the Sec. 4.1 heading 'Then obs-channel 1-Dimensional CNN' should likely read 'The n-obs-channel 1-Dimensional CNN'.
  2. [Table 1] The caption should explicitly state that the mono-Z one-component fermion and scalar mass ranges are disjoint, since this is crucial for interpreting the spin-classification result.
  3. [Sec. 3.1.2] After promoting the DM fields to SU(2)_L doublets, referring to Eq. (7) (written for singlet fields) is confusing. The doublet Lagrangian should be written out explicitly or the notation should be clarified.
  4. [Fig. 5 caption] The phrase 'the bin that encompasses the masses between 100 and 200 GeV was suppressed from the plot for visibility' is unclear. Please state what is suppressed, why, and how the reader should interpret the plot.
  5. [Sec. 4.3.1] The statement that for mono-Z scalar DM 'the network predicts the mass with no deviation for nearly 85% of the tested events' must mean 'within some small tolerance'; the wording should be corrected to avoid implying exact predictions.

Circularity Check

1 steps flagged

Mono-Z spin classification is confounded with disjoint one-component mass ranges, so the spin-extraction claim reduces to mass-threshold classification by construction.

specific steps
  1. other [Table 1 (Section 3.1.2); Section 4.2]
    "Table 1 ... Mono-Z ... 1DM mχ ∈ [230, 1000] GeV; ms ∈ [80,230] GeV ... Notice that, for the mono-Z signature, the mass ranges differ between the one- and two-component DM cases. ... When dealing with mono-Z events, it is expected for the network to be able to distinguish between the spins of the particles in the one-component DM case, given the angular distance between the two leptons in the final state."

    The one-component fermion and scalar mono-Z samples are defined on disjoint pole-mass intervals (fermions ≥230 GeV, scalars ≤230 GeV). Thus the spin label is perfectly determined by the mass label in the training data. The CNN's mono-Z spin discrimination accuracy can be reproduced by any mass-correlated kinematic threshold (e.g., on pT or MET); the reported performance does not isolate Δφ or any spin-sensitive observable. Because no mass-matched fermion/scalar control is presented, the 'spin extraction' result is an artifact of the training-data construction rather than evidence of spin sensitivity.

full rationale

The paper is otherwise a self-contained supervised-learning benchmark: events are simulated with MadGraph/Pythia/Delphes, regressors and classifiers are evaluated on held-out test sets drawn from the same mass points, and the cross-section balancing condition (Eq. 8) and coupling rescaling (Eq. 10) are openly stated modeling assumptions, not hidden fits to the target claims. Mass regression and one-vs-two-component classification do not reduce by construction. No load-bearing self-citation chain is used. The single significant circularity is the mono-Z spin claim: Table 1 assigns non-overlapping mass ranges to the fermion and scalar one-component samples, making spin and mass labels identical by construction. The classifier therefore need not use any spin-dependent information, and the abstract's claim of extracting spin is not demonstrated. This is partial circularity, not a fully circular derivation.

Axiom & Free-Parameter Ledger

6 free parameters · 5 axioms · 1 invented entities

The central claim rests on a curated dataset: couplings are hand-tuned per mass point to balance cross sections (eq. 8, eq. 10), only normalized shapes are used, and the background-free hypothesis is assumed. The CNN's inferred abilities are therefore conditional on these modeling choices, not on any external benchmark.

free parameters (6)
  • g_S(m_chi) per mass bin (mono-jet) = 7.4e-6 to 2.94 (Table 1)
    Rescaled per mass point via eq. (10) to keep the fermionic DM production cross section nearly mass-independent and to satisfy the comparable-cross-section condition (eq. 8). Determines how visible the fermion is in the histograms.
  • mu_phiSS(m_S) per mass bin (mono-jet) = 4.2e-8 to 1200 (Table 1)
    Varied per mass point to bring the scalar DM cross section into the same order of magnitude as the fermionic one (eq. 8 and Figures 5-6). Central to the balanced dataset.
  • Cross-section ratio parameter r = 1, 2, 3, 5 (main results r=1)
    Defines the two-component DM scenario in eq. (8). The main study uses r=1, meaning the cross sections are required to be within 15% of each other.
  • Fixed couplings for mono-Z (g_S, mu_phiSS) = g_S=0.2, mu_phiSS=125 GeV
    Chosen so that production cross sections in the mono-Z channel depend mostly on the DM masses, allowing mass-pair selection for comparable rates.
  • NN hyperparameters (learning rate, filters, early stopping patience) = learning rate 1e-4 to 1e-3; filters/patience vary
    Tuned per classification and regression task; the architecture skeleton is given in Tables 3-4 but exact values are not published. These affect reported performance.
  • Mixing angle alpha and fixed quartic couplings (lambda, lambda_phi_phiSS) = alpha small (<4-10% from Refs [41-43]); lambda=0.01/0.057, lambda_phi_phiSS=0.02
    Set to satisfy Higgs constraints and used in the scalar potential. They influence the mediator-DM couplings and thus the kinematics.
axioms (5)
  • domain assumption Two-component DM model of Ref. [24] (fermion chi, scalar S, scalar mediator phi, one Z2 symmetry)
    Entire simulation is built on this Lagrangian (eqs. 1-2) and stability assumptions; no alternative dark sector models are tested.
  • domain assumption Signal-only (background-free) hypothesis for both mono-jet and mono-Z
    Stated in Section 4: 'we have assumed the signal-only hypothesis with no background.' This is an idealization; real LHC data has irreducible backgrounds (e.g., Z to nu nu + jets).
  • ad hoc to paper Normalized histogram shapes (not absolute rates) contain sufficient information
    Section 3.2.1: 'we consider only the shapes of the kinematical distributions (normalize all distributions to 1), not their absolute event rates.' This discards cross-section information that would carry mass/coupling dependence.
  • domain assumption Validity of the Monte Carlo toolchain (SARAH/SPHENO/MadGraph/PYTHIA/Delphes) for LO simulation and detector response
    Relied on without cross-checks against NLO or data; the paper uses LO matrix elements and Delphes fast simulation.
  • domain assumption Small mixing angle alpha between Higgs and mediator, satisfying Ref. [41] constraints
    Equation (3) and the following text assume alpha small enough to avoid Higgs constraints; this restricts the model to a narrow parameter region.
invented entities (1)
  • SU(2)_L doublet dark fermion chi and dark scalar S (mono-Z realization) no independent evidence
    purpose: To make the DM candidates couple to the Z boson, enabling mono-Z production with final-state radiation, which the paper argues carries spin information.
    The doublet promotion is a modification of the singlet model of Ref. [24], introduced in Section 2 to enable the mono-Z analysis. No direct collider evidence exists for these states.

pith-pipeline@v1.3.0-daily-deepseek · 17106 in / 20679 out tokens · 199112 ms · 2026-08-05T04:47:46.468274+00:00 · methodology

0 comments
Cite this review

Pith. "Pith review of Mono-X Signal Characterization from Two-component Dark Matter Using a Convolutional Neural Network." pith.science (2026). https://pith.science/paper/4TH4KP3J

@misc{pith2026260803975,
  author       = {Pith},
  title        = {Pith review of: Mono-X Signal Characterization from Two-component Dark Matter Using a Convolutional Neural Network},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4TH4KP3J}},
  note         = {Machine review of arXiv:2608.03975}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

We assess the scope of a Convolutional Neural Network (CNN) in characterizing potential signals of two-component Dark Matter (DM) arising at the Large Hadron Collider (LHC) from mono-jet and mono-Z probes. We show that such a CNN has the ability of not only inferring the presence of two DM particles but also of extracting their mass and spin, the latter being either 0 or 1/2, following detector level analysis. However, such result represents a conceptual proof-of-concept, as we have not entertained a signal-to-background analysis.

Figures

Figures reproduced from arXiv: 2608.03975 by Max Fust\'e Costa, Stefano Moretti, Yong Sheng Koay.

Figure 1
Figure 1. Figure 1: Example of topology giving rise to a mono-jet signature in a proton-proton collision into a DM pair, [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Generic topology giving rise to a mono-Z signature in a proton-proton collision into a DM pair, with a Z boson (decaying into a lepton pair, l = e,µ) arising from (potentially) anywhere in the event. produced through final state radiation [44]. In order to achieve this, we need to make the DM candidates charged under the SU(2)L group. The simplest way to do so is to promote the DM candidates to be SU(2)L d… view at source ↗
Figure 3
Figure 3. Figure 3: Coupling between the DM fermion χ and the scalar mediator φ. The coupling constant related to this interaction is gS . φ S S φ φ S S S S H [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Couplings between the DM scalar S and the mediator φ (left and centre) as well as the SM Higgs state (right). The associated constants are, from left to right, µφSS , λφφSS and λ. Once φ acquires a VEV, the second diagram also contributes to the three-point interaction φSS. The coupling cannot be arbitrarily large or the theory becomes non-perturbative. As the perturbative limit we choose gS < 3. (9) As we… view at source ↗
Figure 5
Figure 5. Figure 5: Rescaling of the coupling constants for the mono-jet case. The orange curve, dotted, represents the [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Phase space of the masses of the fermionic and scalar component that fulfill all the given condi [PITH_FULL_IMAGE:figures/full_fig_p008_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Channels that produce a pair of DM particles via a mono- [PITH_FULL_IMAGE:figures/full_fig_p009_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Studied distributions for the mono-jet signature for each of the different signals. From left to right, [PITH_FULL_IMAGE:figures/full_fig_p010_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Studied distributions for the mono-Z signature in the case where only one DM particle (here, the DM fermion, χ) is produced. From left to right, they are pT and MET (top) plus η and ∆φ (bottom). through ∆φ [PITH_FULL_IMAGE:figures/full_fig_p011_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Visual representation of the workflow for the mono-jet signature with 1 ML classifier and 2 ML [PITH_FULL_IMAGE:figures/full_fig_p012_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Visual representation of the workflow for the mono- [PITH_FULL_IMAGE:figures/full_fig_p012_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Schematic illustration of an n-1DCNN with 3 observables. [PITH_FULL_IMAGE:figures/full_fig_p013_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Confusion matrix for mono-jet (left) and mono- [PITH_FULL_IMAGE:figures/full_fig_p015_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: ROC curve for NN classification of DM signals for the mono-jet data (left) and mono- [PITH_FULL_IMAGE:figures/full_fig_p015_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: Relative error in the mass prediction for both DM particles in the one-component DM case for the [PITH_FULL_IMAGE:figures/full_fig_p016_15.png] view at source ↗
Figure 16
Figure 16. Figure 16: Standard deviation of the relative error, [PITH_FULL_IMAGE:figures/full_fig_p017_16.png] view at source ↗
Figure 17
Figure 17. Figure 17: Fraction of events with an absolute mass deviation below a certain threshold in the one-component [PITH_FULL_IMAGE:figures/full_fig_p017_17.png] view at source ↗
Figure 18
Figure 18. Figure 18: Relative error in the mass prediction for both DM particles in the one-component DM case for the [PITH_FULL_IMAGE:figures/full_fig_p018_18.png] view at source ↗
Figure 19
Figure 19. Figure 19: Fraction of events with an absolute mass deviation below a certain threshold in the one-component [PITH_FULL_IMAGE:figures/full_fig_p018_19.png] view at source ↗
Figure 20
Figure 20. Figure 20: Relative error in the mass prediction for both DM particles in the two-component DM case for the [PITH_FULL_IMAGE:figures/full_fig_p019_20.png] view at source ↗
Figure 21
Figure 21. Figure 21: Standard deviation of the relative error [PITH_FULL_IMAGE:figures/full_fig_p019_21.png] view at source ↗
Figure 22
Figure 22. Figure 22: Fraction of events with an absolute mass deviation below a certain threshold in the two-component [PITH_FULL_IMAGE:figures/full_fig_p019_22.png] view at source ↗
Figure 23
Figure 23. Figure 23: Relative error in the mass prediction for both DM particles in the two-component DM case for the [PITH_FULL_IMAGE:figures/full_fig_p020_23.png] view at source ↗
Figure 24
Figure 24. Figure 24: Standard deviation of the relative error, [PITH_FULL_IMAGE:figures/full_fig_p020_24.png] view at source ↗
Figure 25
Figure 25. Figure 25: Fraction of events with an absolute mass deviation below a certain threshold in the two-component [PITH_FULL_IMAGE:figures/full_fig_p021_25.png] 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

62 extracted references · 22 canonical work pages · 17 internal anchors

  1. [1]

    Zwicky, Helv

    F . Zwicky, Helv. Phys. Acta6(1933), 110-127

    work page 1933

  2. [2]

    V . C. Rubin and W . K. Ford, Jr., Astrophys. J.159(1970), 379-403

    work page 1970

  3. [3]

    Corbelli and P

    E. Corbelli and P . Salucci, Mon. Not. Roy. Astron. Soc.311(2000), 441-447 [arXiv:astro-ph/9909252 [astro- ph]]

    Pith/arXiv arXiv 2000

  4. [4]

    Clowe, M

    D. Clowe, M. Bradac, A. H. Gonzalez, M. Markevitch, S. W . Randall, C. Jones and D. Zaritsky, Astrophys. J. Lett.648(2006), L109-L113 [arXiv:astro-ph/0608407 [astro-ph]]

    Pith/arXiv arXiv 2006

  5. [5]

    Aghanimet al.[Planck], Astron

    N. Aghanimet al.[Planck], Astron. Astrophys.641(2020), A6 [erratum: Astron. Astrophys.652(2021), C4] [arXiv:1807.06209 [astro-ph.CO]]

    Pith/arXiv arXiv 2020

  6. [6]

    Bertone, D

    G. Bertone, D. Hooper and J. Silk, Phys. Rept.405(2005), 279-390. [arXiv:hep-ph/0404175 [hep-ph]]

    Pith/arXiv arXiv 2005

  7. [7]

    Buchmueller, C

    O. Buchmueller, C. Doglioni and L. T . Wang, Nature Phys.13(2017) no.3, 217-223 [arXiv:1912.12739 [hep- ex]]

    Pith/arXiv arXiv 2017

  8. [8]

    Dunjko and H

    V . Dunjko and H. J. Briegel, Rept. Prog. Phys.81(2018) no.7, 074001

    work page 2018

  9. [9]

    Krenn, M

    M. Krenn, M. Malik, R. Fickler, R. Lapkiewicz and A. Zeilinger, Phys. Rev. Lett.116(2016) no.9, 090405

    work page 2016

  10. [10]

    Koch-Janusz and Z

    M. Koch-Janusz and Z. Ringel, Nature Phys.14(2018) no.6, 578-582 [arXiv:1704.06279 [cond-mat.dis-nn]]

    Pith/arXiv arXiv 2018

  11. [11]

    Feickert and B

    M. Feickert and B. Nachman, [arXiv:2102.02770 [hep-ph]]

    Pith/arXiv arXiv

  12. [12]

    A. J. Brennan, M. F . McDonald, J. Gramling and T . D. Jacques, JHEP05(2016), 112 [arXiv:1603.01366 [hep- ph]]

    Pith/arXiv arXiv 2016

  13. [13]

    S. P . Liew, M. Papucci, A. Vichi and K. M. Zurek, JHEP06(2017), 082 [arXiv:1612.00219 [hep-ph]]

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  14. [14]

    Actual Physics behind Mono-X

    E. Bernreuther, J. Horak, T . Plehn and A. Butter, SciPost Phys.5(2018) no.4, 034 [arXiv:1805.11637 [hep- ph]]

    work page internal anchor Pith review Pith/arXiv arXiv 2018

  15. [15]

    Cranmer, J

    K. Cranmer, J. Brehmer and G. Louppe, Proc. Nat. Acad. Sci.117(2020) no.48, 30055-30062 [arXiv:1911.01429 [stat.ML]]

    Pith/arXiv arXiv 2020

  16. [16]

    Brehmer and K

    J. Brehmer and K. Cranmer, [arXiv:2010.06439 [hep-ph]]. 20

    Pith/arXiv arXiv 2010

  17. [17]

    C. K. Khosa, V . Sanz and M. Soughton, SciPost Phys.10(2021) no.6, 151 [arXiv:1910.06058 [hep-ph]]

    work page internal anchor Pith review Pith/arXiv arXiv 2021

  18. [18]

    Arganda, A

    E. Arganda, A. D. Medina, A. D. Perez and A. Szynkman, SciPost Phys.12(2022) no.2, 063 [arXiv:2105.12018 [hep-ph]]

    Pith/arXiv arXiv 2022

  19. [19]

    Celik, Eur

    A. Celik, Eur. Phys. J. C83, no.12, 1150 (2023)

    work page 2023

  20. [20]

    H. Lv, D. Wang and L. Wu, Phys. Rev. D106, no.5, 055008 (2022) [arXiv:2203.14569 [hep-ph]]

    work page internal anchor Pith review Pith/arXiv arXiv 2022

  21. [21]

    Boosting mono-jet searches with model-agnostic machine learning

    T . Finke, M. Krämer, M. Lipp and A. Mück, JHEP08, 015 (2022) [arXiv:2204.11889 [hep-ph]]

    work page internal anchor Pith review Pith/arXiv arXiv 2022

  22. [22]

    Gianotti, M

    F . Gianotti, M. L. Mangano, T . Virdee, S. Abdullin, G. Azuelos, A. Ball, D. Barberis, A. Belyaev, P . Bloch and M. Bosman,et al.Eur. Phys. J. C39, 293-333 (2005) [arXiv:hep-ph/0204087 [hep-ph]]

    Pith/arXiv arXiv 2005

  23. [23]

    Bhattacharya, A

    S. Bhattacharya, A. Drozd, B. Grzadkowski and J. Wudka, JHEP10(2013), 158 [arXiv:1309.2986 [hep-ph]]

    Pith/arXiv arXiv 2013

  24. [24]

    S. Esch, M. Klasen and C. E. Yaguna, JHEP09(2014), 108 [arXiv:1406.0617 [hep-ph]]

    Pith/arXiv arXiv 2014

  25. [25]

    Aoki and T

    M. Aoki and T . Toma, JCAP01(2017), 042 [arXiv:1611.06746 [hep-ph]]

    Pith/arXiv arXiv 2017

  26. [26]

    A Minimal Model For Two-Component FIMP Dark Matter: A Basic Search

    S. Peyman Zakeri, S. Mohammad Moosavi Nejad, M. Zakeri and S. Yaser Ayazi, Chin. Phys. C42(2018) no.7, 073101 [arXiv:1801.09115 [hep-ph]]

    work page internal anchor Pith review Pith/arXiv arXiv 2018

  27. [27]

    Two-component dark matter and a massless neutrino in a new B-L model

    N. Bernal, D. Restrepo, C. Yaguna and Ó. Zapata, Phys. Rev. D99(2019) no.1, 015038 [arXiv:1808.03352 [hep-ph]]

    work page internal anchor Pith review Pith/arXiv arXiv 2019

  28. [28]

    Minimal Two-component Scalar Doublet Dark Matter with Radiative Neutrino Mass

    D. Borah, R. Roshan and A. Sil, Phys. Rev. D100(2019) no.5, 055027 [arXiv:1904.04837 [hep-ph]]

    work page internal anchor Pith review Pith/arXiv arXiv 2019

  29. [29]

    Aranda, D

    A. Aranda, D. Hernández-Otero, J. Hernández-Sanchez, V . Keus, S. Moretti, D. Rojas-Ciofalo and T . Shin- dou, Phys. Rev. D103, no.1, 015023 (2021) [arXiv:1907.12470 [hep-ph]]

    Pith/arXiv arXiv 2021

  30. [30]

    The $Z_5$ model of two-component dark matter

    G. Bélanger, A. Pukhov, C. E. Yaguna and Ó. Zapata, JHEP09(2020), 030 [arXiv:2006.14922 [hep-ph]]

    work page internal anchor Pith review Pith/arXiv arXiv 2020

  31. [31]

    Multi-component Dark Matter in a Simplified E$_6$SSM Model

    S. Khalil, S. Moretti, D. Rojas-Ciofalo and H. Waltari, Phys. Rev. D102(2020) no.7, 075039 [arXiv:2007.10966 [hep-ph]]

    work page internal anchor Pith review Pith/arXiv arXiv 2020

  32. [32]

    Two Component Doublet-Triplet Scalar Dark Matter stabilising the Electroweak vacuum

    N. Chakrabarty, R. Roshan and A. Sil, Phys. Rev. D105(2022) no.11, 115010 [arXiv:2102.06032 [hep-ph]]

    work page internal anchor Pith review Pith/arXiv arXiv 2022

  33. [33]

    C. E. Yaguna and Ó. Zapata, Phys. Rev. D105(2022) no.9, 095026 [arXiv:2112.07020 [hep-ph]]

    work page internal anchor Pith review Pith/arXiv arXiv 2022

  34. [34]

    The $Z_3$ soft breaking in the I(2+1)HDM and its probes at present and future colliders

    D. Hernández-Otero, J. Hernández-Sánchez, S. Moretti and T . Shindou, [arXiv:2203.06323 [hep-ph]]

    work page internal anchor Pith review Pith/arXiv arXiv

  35. [35]

    Hernandez-Sanchez, V

    J. Hernandez-Sanchez, V . Keus, S. Moretti, D. Rojas-Ciofalo and D. Sokolowska, [arXiv:2012.11621 [hep- ph]]

    Pith/arXiv arXiv 2012

  36. [36]

    Hernandez-Sanchez, V

    J. Hernandez-Sanchez, V . Keus, S. Moretti and D. Sokolowska, JHEP03, 045 (2023) [arXiv:2202.10514 [hep- ph]]

    Pith/arXiv arXiv 2023

  37. [37]

    A combined approach to the analysis of space and ground experimental data within a simplified E6SSM

    S. Khalil, K. Kowalska, S. Moretti, D. Rojas-Ciofalo and H. Waltari, Eur. Phys. J. C82, no.11, 1058 (2022) [arXiv:2107.10780 [hep-ph]]

    work page internal anchor Pith review Pith/arXiv arXiv 2022

  38. [38]

    M. Aoki, M. Duerr, J. Kubo and H. Takano, Phys. Rev. D86(2012), 076015 [arXiv:1207.3318 [hep-ph]]

    work page internal anchor Pith review Pith/arXiv arXiv 2012

  39. [39]

    C. E. Yaguna and Ó. Zapata, JHEP03(2020), 109 [arXiv:1911.05515 [hep-ph]]

    work page internal anchor Pith review Pith/arXiv arXiv 2020

  40. [40]

    Bélanger, A

    G. Bélanger, A. Pukhov, C. E. Yaguna and Ó. Zapata, JHEP03(2023), 100 [arXiv:2212.07488 [hep-ph]]

    Pith/arXiv arXiv 2023

  41. [41]

    Barateet al.[LEP Working Group for Higgs boson searches, ALEPH, DELPHI, L3 and OPAL], Phys

    R. Barateet al.[LEP Working Group for Higgs boson searches, ALEPH, DELPHI, L3 and OPAL], Phys. Lett. B565(2003), 61-75 [arXiv:hep-ex/0306033 [hep-ex]]

    Pith/arXiv arXiv 2003

  42. [42]

    Aadet al.[ATLAS], Phys

    G. Aadet al.[ATLAS], Phys. Rev. D101(2020) no.1, 012002 [arXiv:1909.02845 [hep-ex]]

    Pith/arXiv arXiv 2020

  43. [43]

    Tumasyanet al.[CMS], Nature607(2022) no.7917, 60-68 [arXiv:2207.00043 [hep-ex]]

    A. Tumasyanet al.[CMS], Nature607(2022) no.7917, 60-68 [arXiv:2207.00043 [hep-ex]]

    Pith/arXiv arXiv 2022

  44. [44]

    Dark Matter Spin Characterisation in Mono-$Z$ Channels

    W . Abdallah, A. Hammad, S. Khalil and S. Moretti, Phys. Rev. D100(2019) no.9, 095006 [arXiv:1907.08358 [hep-ph]]

    work page internal anchor Pith review Pith/arXiv arXiv 2019

  45. [45]

    Staub, [arXiv:0806.0538 [hep-ph]]

    F . Staub, [arXiv:0806.0538 [hep-ph]]. 21

    Pith/arXiv arXiv

  46. [46]

    Staub, Comput

    F . Staub, Comput. Phys. Commun.184(2013), 1792-1809 [arXiv:1207.0906 [hep-ph]]

    Pith/arXiv arXiv 2013

  47. [47]

    Staub, Comput

    F . Staub, Comput. Phys. Commun.185(2014), 1773-1790 [arXiv:1309.7223 [hep-ph]]

    Pith/arXiv arXiv 2014

  48. [48]

    Porod, Comput

    W . Porod, Comput. Phys. Commun.153(2003), 275-315 [arXiv:hep-ph/0301101 [hep-ph]]

    Pith/arXiv arXiv 2003

  49. [49]

    Porod and F

    W . Porod and F . Staub, Comput. Phys. Commun.183(2012), 2458-2469 [arXiv:1104.1573 [hep-ph]]

    Pith/arXiv arXiv 2012

  50. [50]

    Alwall, R

    J. Alwall, R. Frederix, S. Frixione, V . Hirschi, F . Maltoni, O. Mattelaer, H. S. Shao, T . Stelzer, P . Torrielli and M. Zaro, JHEP07, 079 (2014) [arXiv:1405.0301 [hep-ph]]

    Pith/arXiv arXiv 2014

  51. [51]

    Sjostrand, S

    T . Sjostrand, S. Mrenna and P . Z. Skands, Comput. Phys. Commun.178(2008), 852-867 [arXiv:0710.3820 [hep-ph]]

    Pith/arXiv arXiv 2008

  52. [52]

    Sjöstrand, S

    T . Sjöstrand, S. Ask, J. R. Christiansen, R. Corke, N. Desai, P . Ilten, S. Mrenna, S. Prestel, C. O. Rasmussen and P . Z. Skands, Comput. Phys. Commun.191(2015), 159-177 [arXiv:1410.3012 [hep-ph]]

    Pith/arXiv arXiv 2015

  53. [53]

    de Favereauet al.[DELPHES 3], JHEP02(2014), 057 [arXiv:1307.6346 [hep-ex]]

    J. de Favereauet al.[DELPHES 3], JHEP02(2014), 057 [arXiv:1307.6346 [hep-ex]]

    Pith/arXiv arXiv 2014

  54. [54]

    Conte, B

    E. Conte, B. Fuks and G. Serret, Comput. Phys. Commun.184, 222-256 (2013) [arXiv:1206.1599 [hep-ph]]

    Pith/arXiv arXiv 2013

  55. [55]

    Brehmer, G

    J. Brehmer, G. Louppe, J. Pavez and K. Cranmer, Proc. Nat. Acad. Sci.117(2020) no.10, 5242-5249 [arXiv:1805.12244 [stat.ML]]

    Pith/arXiv arXiv 2020

  56. [56]

    Hallin, J

    A. Hallin, J. Isaacson, G. Kasieczka, C. Krause, B. Nachman, T . Quadfasel, M. Schlaffer, D. Shih and M. Som- merhalder, Phys. Rev. D106(2022) no.5, 055006 [arXiv:2109.00546 [hep-ph]]

    Pith/arXiv arXiv 2022

  57. [57]

    Nachman and J

    B. Nachman and J. Thaler, Phys. Rev. D103(2021) no.11, 116013 [arXiv:2101.07263 [physics.data-an]]

    Pith/arXiv arXiv 2021

  58. [58]

    Heinrich, S

    L. Heinrich, S. Mishra-Sharma, C. Pollard and P . Windischhofer, [arXiv:2306.12584 [stat.ML]]

    Pith/arXiv arXiv

  59. [59]

    E. M. Metodiev, B. Nachman and J. Thaler, JHEP10(2017), 174 [arXiv:1708.02949 [hep-ph]]

    Pith/arXiv arXiv 2017

  60. [60]

    Nachman and D

    B. Nachman and D. Shih, Phys. Rev. D101(2020), 075042 [arXiv:2001.04990 [hep-ph]]

    Pith/arXiv arXiv 2020

  61. [61]

    Papamakarios, E

    G. Papamakarios, E. Nalisnick, D. J. Rezende, S. Mohamed and B. Lakshminarayanan, J. Machine Learning Res.22(2021) no.1, 2617-2680 [arXiv:1912.02762 [stat.ML]]

    Pith/arXiv arXiv 2021

  62. [62]

    Papamakarios, T

    G. Papamakarios, T . Pavlakou and I. Murray, [arXiv:1705.07057 [stat.ML]]. 22

    Pith/arXiv arXiv