{"id":"ae76a4c1-9e2c-4f5f-bd1d-f503e479a490","arxiv_id":"1908.01713","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":5,"one_line_summary":"A CMS search combining five Higgs decay channels finds no dark matter signal and sets limits on mono-Higgs production in two simplified models.","lead":"CMS searched for dark matter particles produced together with a Higgs boson in 13 TeV proton-proton collisions, using five Higgs decay channels. No signal was found, and the result sets new exclusion limits on two simplified dark matter models.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified","rationale":"The reader's verdict of ACCEPT is appropriate. The paper is a well-executed experimental search with a clear null result. The weakest assumption identified by the reader—unbiased background predictions and transfer factors, especially for nonprompt leptons in the new channels—is a legitimate point of caution, but it does not rise to the level of a load-bearing concern because those channels contribute only marginally to the combined limit. The h→bb channel dominates the combination, and its methods are documented in the referenced publications [30,31]. The paper is notably transparent about the Z'-2HDM benchmark having been excluded by dijet searches, which is a serious limitation for the model-specific interpretation but is explicitly acknowledged and does not undermine the primary experimental statement. The only internal inconsistency found is the mχ typo in Section 8, which is inconsequential. Overall, the evidence supporting the central claim is strong within the scope of a text-based review, and no adjustment to the verdict is needed.","tokens_in":49718,"tokens_out":12449,"duration_ms":128272,"concrete_test":"Refit the combined 95% CL exclusion without the h→WW and h→ZZ channels and compare the observed mZ' boundaries with the full combination; if the boundaries shift by less than the expected ±1σ band for both the Z'-2HDM and baryonic Z' models, the reader's weakest assumption (nonprompt-lepton background systematics in the new channels) is confirmed to be non-load-bearing.","verdict_should_be":"UNCHANGED","load_bearing_attack":"After a careful second pass, I do not find a load-bearing concern that would invalidate the central claim. The reader's identified weakness—the data-driven nonprompt-lepton background in the new WW and ZZ channels, with 30% and 36–43% uncertainties (Sections 5.4.1 and 5.5.1, Tables 3 and 4)—is genuine but is not load-bearing for the combined result. These channels contribute only about 660 of 11030 expected background events in WW and 16.2 of 117.8 in ZZ, and their expected signal yields are small; the combined sensitivity is dominated by the h→bb channel. A systematic shift in control-region composition would therefore shift the combined limit by much less than the quoted uncertainty. The paper's self-admitted statement (Section 1) that the Z'-2HDM benchmark mZ' values within sensitivity reach are already excluded by dijet searches is an honest and clearly flagged limitation; it reduces the phenomenological relevance of the Z'-2HDM limits but does not make the null result or the derived limits incorrect. The mχ text inconsistency in Section 8 (which says 500 and 1000 GeV while Table 6 and the limits use mχ = 1 GeV) is a typographical error with no effect on the analysis. The central claim is supported by the data, the systematic uncertainties are described in detail, and the combination procedure is internally consistent.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper reports a CMS search for dark matter produced in association with a Higgs boson (mono-Higgs) using 35.9 fb^-1 of 13 TeV proton-proton collisions. The analysis covers five Higgs decay channels (bb, gamma gamma, tau tau, WW, and ZZ); the WW and ZZ channels are new to this search, and the five channels are statistically combined. The central result is a null observation: data agree with the SM background expectation in all channels and in the combination, and 95% CL upper limits are set on the Z'-2HDM and baryonic Z' simplified models, including an interpretation as a spin-independent DM-nucleon scattering cross section. The paper claims the first mono-Higgs search in the WW and ZZ final states and the first five-channel mono-Higgs combination.","tokens_in":50004,"tokens_out":5318,"duration_ms":54659,"significance":"The result, if correct, is a valid upper-limit measurement with a modest but real extension of the mono-Higgs program: the WW and ZZ channels are new, and the five-channel combination improves coverage at low and high mZ'. I was unable to find a load-bearing technical flaw. The analysis uses standard CMS object reconstruction, data-driven control regions for the dominant backgrounds, and a detailed systematic budget; the observed yields agree with the expected background within uncertainties. The paper is also commendable for explicitly stating in Section 1 that the Z'-2HDM benchmark points within sensitivity reach are already excluded by dijet searches; this limits the phenomenological priority of the Z'-2HDM limits but does not make the null result or the derived limits incorrect. The large nonprompt-lepton uncertainties in the new WW and ZZ channels (30% and 36-43%, Tables 3 and 4) are real but not decisive: the skeptic's stress-test concern does not land, because those channels contribute only a small fraction of the combined expected signal and background, and the combination is dominated by h->bb.","major_comments":[],"minor_comments":[{"comment":"The text says the baryonic Z' signal benchmark is normalized to mZ' = 500 GeV and mchi = 1000 GeV, but Table 6, Table 7, and all the figures and limits in the paper use mchi = 1 GeV for this benchmark; this is an obvious typo and should be corrected.","section":"Section 8, text before Tables 6 and 7"},{"comment":"The axis labels and legend entries in Fig. 2 are garbled in the manuscript (for example, entries like ' = 300 GeV A = 600 GeV, mZ'm = 400 GeV' should read something like 'mA = 300 GeV, mZ' = 400 GeV'); please restore the full parameter notation and verify every legend entry.","section":"Figure 2 caption"},{"comment":"The caveat that the chosen Z'-2HDM benchmark has mZ' values within sensitivity reach already excluded by dijet searches is stated clearly in the introduction; I recommend repeating this caveat in the summary section so that the final conclusions do not overstate the reach of the Z'-2HDM limits.","section":"Section 1 and Section 9"},{"comment":"The description of the BDT training is ambiguous: the sentence 'simulated signal samples with mA = 300 GeV (mchi = 1 GeV) with various values of mZ' have been used for training' could mean either a single training sample per model or an ensemble of samples; please clarify which signal points enter the training and whether the same BDT is applied to all scan points.","section":"Section 5.4.1"},{"comment":"The observed yields in the WW and ZZ channels are close to the predicted backgrounds (11,172 vs. 11,030 and 112 vs. 117.8, respectively); a sentence reporting the local p-value or significance of the largest deviation would make the 'no significant excess' statement more quantitative.","section":"Section 8.1, Tables 6 and 7"}],"recommendation":"minor_revision","confidential_remarks":"I see no grounds for rejection. The only substantive caveat is the self-flagged dijet exclusion of the Z'-2HDM benchmark, which reduces the physics impact of that half of the paper but does not compromise correctness. The mchi typo in Section 8 and the garbled Fig. 2 caption should be corrected before publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is a well-executed CMS search that does what it claims — no excess, limits in two simplified models — and the new pieces are real: first mono-Higgs results in h→WW and h→ZZ, and a five-channel combination. It deserves a serious referee.\n\nWhat's good: The analysis uses standard CMS objects, data-driven control regions for the tricky nonprompt-lepton backgrounds, and a careful correlation treatment in the combination. The h→bb channel dominates sensitivity; the WW and ZZ channels are new but small contributors. The paper is unusually honest about the Z'-2HDM benchmark: Section 1 states that the mZ' values within reach are already excluded by ATLAS/CMS dijet searches, and that they keep the benchmark for comparability. That is the right call, and it is a limitation, not a flaw. The baryonic Z' interpretation and direct-detection comparison are useful.\n\nSoft spots, in proportion: The mχ inconsistency in Section 8 is a typo — text says the baryonic Z' signal benchmark uses mχ=500 and 1000 GeV while Table 6 and all limits use 1 GeV. Annoying but harmless. The nonprompt-lepton background uncertainties are 30% in WW and 36–43% in ZZ, but those channels contribute ~660/11030 and ~16/118 expected background events, so the combined limit is insensitive to them. The Z'-2HDM benchmark's prior dijet exclusion reduces phenomenological punch but is not an analysis error.\n\nThe physics reach is modest — 35.9 fb−1, 2016 data only — and the exclusion in Z'-2HDM is largely of already-excluded parameter space. Still, the null result is a valid data point, and the five-channel combination procedure is a useful template.\n\nWho it's for: experimental colleagues working on mono-X or Higgs-portal DM searches, and phenomenologists who need to quote LHC limits. It is not a paper that changes the field; it extends a program carefully.\n\nRecommendation: send it to peer review. It is internally consistent, the systematics are documented, and the honest caveat about the benchmark is exactly the kind of self-reporting that should be encouraged. I would accept after minor revision (fix the mχ text, maybe tighten the benchmark discussion).","headline":"A solid, honestly-reported null search: first mono-Higgs in WW/ZZ plus a five-channel combination, with no serious flaw beyond a benchmark caveat the authors themselves flag.","tokens_in":50507,"tokens_out":1865,"would_cite":true,"duration_ms":18045,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A search for dark matter produced with a Higgs boson in five decay channels finds no excess and sets new 95% CL limits on two simplified mediator models.","keywords":["dark matter","mono-Higgs","missing transverse momentum","Higgs boson","Z' mediator","two-Higgs-doublet model","baryonic Z' model","LHC physics"],"falsifier":"Run the same five-channel combination on an independent dataset of comparable size: if the observed number of events in the high-missing-transverse-momentum signal regions exceeds the background-only prediction by a 5-sigma fluctuation consistent with a Z'->Higgs-plus-dark-matter hypothesis, the no-excess claim would be refuted. A specific place to look is the h->bb invariant-mass sidebands combined with the high-pTmiss tail, where a resonant signal would create a localized excess.","tokens_in":49524,"feed_emoji":"⚛️","tokens_out":4222,"duration_ms":45938,"temperature":0.7,"pith_summary":"This paper searches for dark matter particles produced together with a Higgs boson, using proton-proton collisions at 13 TeV and combining five Higgs decay channels: bottom quarks, photons, tau leptons, W bosons, and Z bosons. No significant excess over the expected standard model background is observed in any channel or in the combined analysis. The study sets upper limits at 95% confidence level on dark matter production in a Z'-extended two-Higgs-doublet model and in a baryonic Z' model, and interprets the results as limits on the spin-independent dark-matter-nucleon scattering cross section. It is the first search for dark matter with a Higgs boson decaying to W or Z pairs and the first statistical combination of all five channels.","feed_headline":"No dark matter seen in Higgs-plus-missing-energy events","feed_subtitle":"First five-channel combination excludes mediator masses up to 3.2 TeV and tightens light-dark-matter limits.","key_machinery":"The central object is the mono-Higgs signature: a Higgs boson recoiling against a pair of invisible dark matter particles, inferred experimentally as a reconstructed Higgs candidate accompanied by large missing transverse momentum. Two simplified benchmark models carry the interpretation: the Z'-2HDM, where a Z' boson decays to a Higgs boson and a pseudoscalar mediator that decays to dark matter, and the baryonic Z' model, where a Z' mediator radiates a Higgs boson before decaying to dark matter. The analysis machinery combines per-channel discriminants—dijet and diphoton invariant masses, tau-pair transverse mass, a boosted decision tree for WW, and four-lepton mass plus missing momentum for ZZ—into a single profile-likelihood fit across all five channels.","core_discovery":"The authors establish that, in 35.9 inverse femtobarns of 13 TeV collision data, events containing a Higgs boson candidate plus large missing transverse momentum are consistent with standard model expectations. In the Z'-2HDM benchmark, the combined analysis excludes Z' mediator masses from roughly 500 to 3200 GeV for a pseudoscalar mass of 300 GeV; in the baryonic Z' model it excludes Z' masses from 100 to 1600 GeV for a dark matter mass of 1 GeV. These limits are interpreted as a spin-independent dark-matter-nucleon cross-section bound that is more stringent than direct-detection limits for dark matter masses between 1 and 5 GeV. The new WW and ZZ channels, while individually less sensitive, add coverage in kinematically distinct regions and are combined with the previously published bb, gamma-gamma, and tau-tau channels.","pith_inferences":["Editorial inference: the large uncertainties in the data-driven nonprompt-lepton background estimates in the WW and ZZ channels (about 30% and 36-43%, respectively) mean these new channels currently add more breadth than raw precision; with more data or better background control they could become substantially more powerful.","Editorial inference: the same combined-analysis framework could be adapted to other invisible-Higgs topologies, such as Higgs production with a Z boson decaying invisibly, by exchanging the signal model while keeping the per-channel discriminants and likelihood structure.","Editorial inference: a natural testable extension is to rerun this analysis on the full Run 2 dataset (about 140 inverse femtobarns), which would roughly double the sensitivity to the mediator masses and sharpen the low-mass dark matter exclusion."],"forward_implications":["If the limits are correct, dark matter particles lighter than about 5 GeV with spin-independent scattering cross sections above roughly 10^-40 cm^2 are excluded by collider data alone.","The h->WW and h->ZZ channels, newly exploited here, provide a route to probe mono-Higgs signals with soft missing-transverse-momentum spectra where the bb channel is less sensitive.","The five-channel combination demonstrates a reusable statistical strategy for Higgs-associated dark matter searches that can be applied to future datasets with higher luminosity.","For the chosen benchmark parameters, the excluded Z' mass ranges overlap with regions probed by dijet resonance searches, so the two search strategies jointly constrain the model parameter space."],"supporting_citations":[{"why":"Supplies the h->bb AK8-jet analysis used for the Z'-2HDM channel in the combination.","marker":"[30]"},{"why":"Supplies the h->bb CA15-jet analysis used for the baryonic Z' channel.","marker":"[31]"},{"why":"Provides the previously published h->gamma-gamma and h->tau-tau analyses that enter the combination.","marker":"[32]"},{"why":"Provides the four-lepton event selection and background estimation strategy adopted for the new h->ZZ channel.","marker":"[42]"},{"why":"Defines the two simplified benchmark signal models that the search interprets.","marker":"[33]"},{"why":"Proposes the mono-Higgs signature as a collider probe of dark matter.","marker":"[22]"},{"why":"Gives the parameter choices and the formula used to convert the results into spin-independent dark-matter-nucleon cross-section limits.","marker":"[37]"}],"fun_headline_variants":["No dark matter in Higgs-plus-missing-energy events","First five-channel Higgs+MET search: null result","CMS excludes dark matter mediators up to 3.2 TeV in Higgs+MET","Dark matter search in Higgs+MET: no excess, tighter limits","Combined Higgs decay channels find no dark matter signal"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The limits assume that the simulated signal acceptance and background predictions, together with the control-region transfer factors used to normalize data, are unbiased within the quoted systematic uncertainties; in the new WW and ZZ channels this rests on data-driven nonprompt-lepton background estimates that carry uncertainties of 30% and 36-43% respectively.","fun_headline_variants_meta":{"raw":{"variants":["No dark matter in Higgs-plus-missing-energy events","First five-channel Higgs+MET search: null result","CMS excludes dark matter mediators up to 3.2 TeV in Higgs+MET","Dark matter search in Higgs+MET: no excess, tighter limits","Combined Higgs decay channels find no dark matter signal"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000162,"raw_usage":{"total_tokens":1244,"prompt_tokens":955,"completion_tokens":289,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":571,"completion_tokens_details":{"reasoning_tokens":203}},"tokens_in":571,"tokens_out":289,"duration_ms":3544,"temperature":1.0,"reasoning_tokens":203,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:05:45.376799+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the same five-channel combination on an independent dataset of comparable size: if the observed number of events in the high-missing-transverse-momentum signal regions exceeds the background-only prediction by a 5-sigma fluctuation consistent with a Z'->Higgs-plus-dark-matter hypothesis, the no-excess claim would be refuted. A specific place to look is the h->bb invariant-mass sidebands combined with the high-pTmiss tail, where a resonant signal would create a localized excess.","supporting_citations":[{"cited_title":"Search for dark matter produced in association with a Higgs boson decaying to a pair of bottom quarks in proton-proton collisions at $\\sqrt{s} =$ 13 TeV","cited_arxiv_id":"1811.06562","evidence_quote":"Supplies the h->bb CA15-jet analysis used for the baryonic Z' channel."}],"review_version":1}