{"id":"87d3cee7-49f2-44d4-bd62-b149d9100208","arxiv_id":"2411.15000","paper_version":2,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"CMS sets 95% confidence upper limits on Z and Higgs decays to J/psi or psi(2S) plus a photon and constrains the ratio of Higgs-charm to Higgs-photon couplings.","lead":"The CMS experiment searched for two ultra-rare decay modes where a Z or Higgs boson decays to a photon plus a J/psi or psi(2S) meson, which then decays to two muons, using 123 inverse femtobarns of LHC proton collisions at 13 TeV. No excess over standard model predictions was seen, and the results set the best or competitive limits in several channels.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified; the data-driven background smoothness assumption is the weakest point, but the discrete profiling method and validation figures render it non-load-bearing.","rationale":"The reader's ACCEPT verdict is well-supported. The strongest claim, that the quoted upper limits are the new CMS reference limits, follows directly from the observed data and the statistical procedure. The weakest assumption identified by the reader, smooth parametric modeling of the nonresonant background, is indeed the most delicate part of the analysis. However, the paper addresses it with a discrete profiling method that includes several function families and a penalty for the number of parameters, and the displayed fits show acceptable agreement. The control region provides an independent handle on the resonant Z Dalitz background, and the asymptotic limits are cross-checked with pseudo-data. The low-statistics categories (VBF, HF) are not the drivers of the final sensitivity. No internal inconsistency, unsupported numerical claim, or circular reasoning was found. The proposed test of adding an external background family would provide a quantitative confirmation that the background-model uncertainty is correctly covered, but the current evidence does not warrant a change in verdict.","tokens_in":37726,"tokens_out":17235,"duration_ms":172909,"concrete_test":"As a stress test, re-derive the background-only and signal-plus-background fits in the four high-statistics signal regions (Z-HP, Z-LP, ggF-HP, ggF-LP) using an additional flexible background family not included in the original discrete profiling set, such as a sum of two exponentials or a fifth-order Bernstein polynomial, and verify that the observed 95% CL limits on B(Z->J/psi gamma) and B(H->J/psi gamma) change by less than the background-model systematic uncertainty assigned in the paper; if the shift is larger, the discrete profiling coverage is insufficient.","verdict_should_be":"UNCHANGED","load_bearing_attack":"I have not identified a load-bearing flaw in the central claim that no evidence is seen and that the quoted 95% CL upper limits are correct. The analysis is careful and standard: the nonresonant QCD background is fitted directly from data in each signal region using power-law, exponential, and Bernstein families, with the discrete profiling method treating the function choice as a discrete nuisance parameter. The control region constrains the Z Dalitz normalization, and the fit quality is documented in Figs. 2-4 with acceptable chi-square values. The 2.2 sigma deficit in the Z->psi(2S) channel is openly reported and is consistent with a statistical fluctuation. The weakest assumption is the smoothness and adequate parametric coverage of the background in m_mu_mu_gamma, but this is a standard assumption in such searches, mitigated by the multiple function families and the discrete profiling penalty. I found no internal inconsistency, missing reference that undermines the result, or unsupported step that would change the central conclusion.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper reports a search for the rare decays Z→J/ψγ, Z→ψ(2S)γ, H→J/ψγ, and H→ψ(2S)γ, using 123 fb−1 of proton-proton collision data at √s = 13 TeV collected by CMS. Events are selected with two opposite-sign muons consistent with a charmonium state and a photon, and are divided into seven signal regions and a control region. The signal is extracted from fits to the m(μμγ) distribution, with the QCD multijet background modeled by parametric functions and the discrete profiling method used to cover the shape uncertainty; electroweak backgrounds are taken from simulation. No significant excess is observed; the most notable feature is a 2.2 standard deviation deficit in the Z→ψ(2S)γ channel. Upper limits at 95% CL are set on the branching fractions: B(Z→J/ψγ) < 0.6×10−6, B(Z→ψ(2S)γ) < 1.3×10−6, B(H→J/ψγ) < 2.6×10−4, and B(H→ψ(2S)γ) < 9.9×10−4. The H→J/ψγ limit is interpreted in the κ framework, giving a constraint on κc/κγ of (−157, +199) at 95% CL, or (−166, +208) assuming κγ = 1.","tokens_in":37839,"tokens_out":12599,"duration_ms":109820,"significance":"If the results are correct, they provide the most stringent limits to date on Z→ψ(2S)γ and H→ψ(2S)γ from CMS and are competitive with the ATLAS 139 fb−1 results for the other two channels. The analysis is careful: the background modeling uses a discrete profiling method, an alternative Voigt signal shape is tested with an effect below 1%, and the CLs limits are cross-checked with pseudo-data, showing agreement within 5%. The interpretation of the H→J/ψγ limit as a constraint on κc/κγ is a valuable addition, as it provides a CMS-based constraint complementary to the ATLAS result. The paper also provides tabulated results in HEPData, which aids reproducibility.","major_comments":[],"minor_comments":[{"comment":"The equation for the ratio µ(H→J/ψγ)/µ(H→γγ) ends with the expression |Aind + Adir κc/κγ|^2 / ΓSM(H→J/ψγ), which is dimensionally opaque; please state the normalization convention for Aind and Adir and note that ΓSM is the width computed from the same amplitudes, so that the expression equals 1 when κc/κγ = 1.","section":"Section 7, Eq. (1)"},{"comment":"The standalone fit to the control region that extracts μ(Z→μμγ) = 1.18 ± 0.12 is reported in the text, but the caption of Fig. 4 states that the resonant background contributions are added with the normalization fixed to the SM expectation; please clarify whether the displayed fit corresponds to the standalone fit or to a fit with fixed resonant normalizations.","section":"Section 7, Fig. 4"},{"comment":"The statement that scale and PDF uncertainties 'are not varied in the fit' when deriving limits on B or σB is ambiguous; in particular, for the branching-fraction limits, which are obtained by assuming the SM production cross section, the paper should state explicitly whether the production cross-section uncertainty is included in the quoted result.","section":"Section 7"},{"comment":"The row 'Resonant bkg. model' lists a 1% normalization uncertainty only for the Z→μμγ background, whereas the text says the alternative Voigtian function is used to cover model biases for both signal and resonant backgrounds; please clarify the extent of this uncertainty.","section":"Section 6, Table 3"},{"comment":"The word 'subsequentially' in the abstract is a typo and should be 'subsequently'.","section":"Abstract and text"},{"comment":"The phrase 'The data set we used does not contain enough events' is informal for a journal article; consider rewording to 'The data set does not contain enough events'.","section":"Section 4"}],"recommendation":"accept","confidential_remarks":"The paper fits the scope of Physics Letters B well: it is a concise search paper with a null result, competitive limits, and a coupling interpretation. I have no concerns about the citation pattern or novelty; the previous CMS and ATLAS searches are properly cited and compared. The statistical treatment is standard and the cross-checks are convincing. My only substantive (but minor) request would be a clarification of the treatment of production cross-section uncertainties in the branching-fraction limits, which could be addressed in a revised version."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First: this is a clean, well-executed null result. It gives the first CMS limits on H/Z to psi(2S)+gamma, improves the H->J/psi gamma limit by roughly a factor of three relative to the 2016 CMS search, and sets a Z->J/psi limit that is a factor of two better than ATLAS's full Run-2 result. The analysis is not trying to oversell: the 2.2 sigma deficit in Z->psi(2S) is stated openly, and the limits are far from the SM predictions, so the paper is honest about being a search in a difficult channel.\n\nWhat it does well: the signal MC now includes spin correlations directly rather than reweighting, the categories for VBF and heavy flavor make sense for the Higgs production modes, and the QCD background is handled with the standard discrete profiling method over multiple function families. The control region constraining the Z Dalitz background is sensible, and the asymptotic CLs limits are cross-checked with pseudo-data within 5%. Systematics are tabulated in detail, including the 1% signal-shape uncertainty from an alternative Voigt function, and the HEPData record is provided. That is the level of reproducibility I expect from a CMS paper, and it is here.\n\nSoft spots are minor. The nonresonant background smoothness in m_mumu_gamma is the main assumption, but the function choices and discrete profiling cover it reasonably, and the validation plots look fine. The VBF category has only 17 events in the displayed fit range, so that fit is not strongly constraining, but it is a small category by design. The kappa_c/kappa_gamma interval is very broad and not a headline result; it is consistent with ATLAS and not surprising given the limit. I do not see a load-bearing flaw. The reader's weakest assumption is the right one, and the stress-test note is right that it does not sink the result.\n\nWho benefits: this is for people working on Higgs and Z rare decays, quarkonium production, and charm Yukawa constraints. It is a reference result for CMS. It deserves a serious referee; I would send it to review with confidence, expecting only clarifications rather than major changes.","headline":"A careful, honest null search with first CMS psi(2S) limits and an improved H->J/psi limit; no load-bearing flaws.","tokens_in":38422,"tokens_out":2691,"would_cite":true,"duration_ms":28469,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"CMS finds no evidence for Z or Higgs boson decays to a J/ψ or ψ(2S) meson and a photon, and sets 95% CL upper limits on the branching fractions.","keywords":["rare decays","Higgs boson","Z boson","J/psi meson","psi(2S) meson","charmonium","branching fraction limits","CMS LHC"],"falsifier":"A re-analysis that finds a localized excess of events at the expected Z or Higgs mass in the m_μμγ distribution, or an alternative background model constructed from a fully data-driven sideband interpolation that yields a significantly different limit, would challenge the result. Concretely, an excess of more than about three standard deviations at m_μμγ ≈ 91 or 125 GeV in the same event selection would contradict the no-excess claim.","tokens_in":37481,"feed_emoji":"⚛️","tokens_out":6540,"duration_ms":54725,"temperature":0.7,"pith_summary":"The paper searches for decays of the standard-model Z and Higgs bosons into a J/ψ or ψ(2S) meson plus a photon, with the charmonium meson decaying to two muons, using 123 fb⁻¹ of 13 TeV proton-proton collisions recorded by CMS. No evidence is seen for any of the four decay channels beyond standard-model expectations, and 95% confidence-level upper limits are placed on the branching fractions. If correct, these are the most sensitive CMS limits for these rare radiative decays to date, and the H→J/ψγ limit translates into a new constraint on the ratio of the Higgs couplings to charm quarks and photons.","feed_headline":"CMS sets limits on rare Z and Higgs decays to charmonium plus a photon","feed_subtitle":"Four 95% CL upper bounds set; the H→J/ψγ limit constrains the Higgs-charm to photon coupling ratio.","key_machinery":"The analysis uses the invariant mass m_μμγ as the final discriminant, with signal modeled by double-sided Crystal Ball functions and the nonresonant QCD background modeled directly from data with power-law, exponential, and Bernstein polynomial functions. The discrete profiling method treats the choice and complexity of the background function as a discrete nuisance parameter, while a likelihood discriminator built from the Z, J/ψ, and muon spin-correlation angles separates Z→J/ψγ signal from background and |cos θ*| separates the gluon-fusion Higgs category.","core_discovery":"The central claim is that the rare decays Z→J/ψγ, Z→ψ(2S)γ, H→J/ψγ, and H→ψ(2S)γ do not appear at rates above the standard-model predictions, and the data are consistent with background only. The observed 95% CL upper limits on the branching fractions are 0.6×10⁻⁶, 1.3×10⁻⁶, 2.6×10⁻⁴, and 9.9×10⁻⁴, respectively, improving on the previous CMS search and comparable to the latest ATLAS results. In the κ-framework interpretation, the H→J/ψγ limit constrains the ratio of Higgs coupling modifiers κ_c/κ_γ to the interval (−157, +199) at 95% CL, or κ_c ∈ (−166, +208) if κ_γ=1 is assumed.","pith_inferences":["Combining these CMS limits with the ATLAS 139 fb⁻¹ results would likely yield even tighter combined constraints on the branching fractions and on κ_c/κ_γ.","The angular-discriminator and discrete-profiling techniques developed here could be transferred to searches for other rare radiative quarkonium decays, such as Z/H→Υ(nS)γ, where the spin correlations differ.","With further LHC data, the Z→J/ψγ search could approach the standard-model prediction near 10⁻⁸, making this channel a plausible future observation target rather than just an exclusion.","If the observed deficit in Z→ψ(2S)γ persists and grows, it would suggest the nonresonant background is overestimated in that region, warranting additional data-driven checks beyond the parametric families used."],"forward_implications":["The new upper limits supersede the previous CMS results from 36 fb⁻¹, improving expected sensitivity by about a factor of two for H→J/ψγ and 30% for Z→J/ψγ through optimized event selection and the discrete profiling treatment of background shapes.","The κ_c/κ_γ constraint provides a probe of the Higgs-charm coupling that is complementary to direct searches for H→cc, and can be combined with those measurements in global fits.","These limits serve as the updated CMS reference for planning future HL-LHC searches for Z→ψ(nS)γ, whose standard-model branching fractions remain roughly two orders of magnitude below current sensitivity.","The 2.2σ downward fluctuation observed in the Z→ψ(2S)γ channel is consistent with background; if data continue to deviate with more statistics, the background model in that channel would merit closer scrutiny."],"supporting_citations":[{"why":"Provides the previous CMS search results at 36 fb⁻¹ that this analysis improves upon.","marker":"[24]"},{"why":"Defines the comparison point for the limits and the methodology for the κ_c/κ_γ interpretation.","marker":"[26]"},{"why":"Supplies the standard-model branching-fraction predictions for Z→J/ψγ and Z→ψ(2S)γ used as signal normalization.","marker":"[18]"},{"why":"Provides the standard-model prediction for H→J/ψγ and the direct and indirect amplitudes used in the κ_c/κ_γ constraint.","marker":"[20]"},{"why":"Describes the discrete profiling method that treats the background-function choice as a discrete nuisance parameter.","marker":"[52]"},{"why":"Supplies the model used to generate H→ψ(nS)γ signal events including spin correlations.","marker":"[17]"},{"why":"Provides the measured ψ(nS)→μμ branching fractions used to convert signal yields to branching-fraction limits.","marker":"[23]"}],"fun_headline_variants":["CMS finds no rare Z or Higgs decays to J/ψ or ψ(2S) plus γ","No signal: CMS sets tight limits on Z, H decays to charmonium+photon","CMS limits on H, Z to charmonium+γ: no excess over SM","Higgs-charm coupling constrained: CMS limits on H→J/ψγ","CMS constrains κ_c/κ_γ to (−157,199) via H→J/ψγ limit"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The claim relies on the assumption that the nonresonant QCD background in each signal region is smooth in m_μμγ and is well described by the power-law, exponential, and Bernstein families considered, so that a narrow signal peak would not be masked by an unmodeled localized background feature.","fun_headline_variants_meta":{"raw":{"variants":["CMS finds no rare Z or Higgs decays to J/ψ or ψ(2S) plus γ","No signal: CMS sets tight limits on Z, H decays to charmonium+photon","CMS limits on H, Z to charmonium+γ: no excess over SM","Higgs-charm coupling constrained: CMS limits on H→J/ψγ","CMS constrains κ_c/κ_γ to (−157,199) via H→J/ψγ limit"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000255,"raw_usage":{"total_tokens":1633,"prompt_tokens":1069,"completion_tokens":564,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":685,"completion_tokens_details":{"reasoning_tokens":443}},"tokens_in":685,"tokens_out":564,"duration_ms":5848,"temperature":1.0,"reasoning_tokens":443,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T14:37:10.759953+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A re-analysis that finds a localized excess of events at the expected Z or Higgs mass in the m_μμγ distribution, or an alternative background model constructed from a fully data-driven sideband interpolation that yields a significantly different limit, would challenge the result. Concretely, an excess of more than about three standard deviations at m_μμγ ≈ 91 or 125 GeV in the same event selection would contradict the no-excess claim.","supporting_citations":[{"cited_title":"Search for rare decays of Z and Higgs bosons to J$/\\psi$ and a photon in proton-proton collisions at $\\sqrt{s} =$ 13 TeV","cited_arxiv_id":"1810.10056","evidence_quote":"Provides the previous CMS search results at 36 fb⁻¹ that this analysis improves upon."},{"cited_title":"Searches for exclusive Higgs and $Z$ boson decays into a vector quarkonium state and a photon using $139$ fb$^{-1}$ of ATLAS $\\sqrt{s}=13$ TeV proton$-$proton collision data","cited_arxiv_id":"2208.03122","evidence_quote":"Defines the comparison point for the limits and the methodology for the κ_c/κ_γ interpretation."},{"cited_title":"Order $v^4$ corrections to Higgs boson decay into $J/\\psi + \\gamma$","cited_arxiv_id":"1907.06473","evidence_quote":"Provides the standard-model prediction for H→J/ψγ and the direct and indirect amplitudes used in the κ_c/κ_γ constraint."},{"cited_title":"Exclusive Radiative Higgs Decays as Probes of Light-Quark Yukawa Couplings","cited_arxiv_id":"1505.03870","evidence_quote":"Supplies the model used to generate H→ψ(nS)γ signal events including spin correlations."}],"review_version":1}