{"id":"49b10780-2fa6-40db-8288-a1db8e532434","arxiv_id":"1908.08500","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"low","formal_verification":"none","parameter_count":3,"one_line_summary":"No supersymmetry signal was found in Higgs-to-diphoton events, and new 95% CL limits exclude sbottom masses below 530 GeV and chargino-neutralino masses below 235-290 GeV in simplified SUSY models.","lead":"The CMS experiment searched 77.5 inverse femtobarns of 13 TeV proton collisions for supersymmetric particles that decay through a Higgs boson into two photons, and found nothing beyond the standard model. The null result tightens previous limits on bottom squarks and on chargino-neutralino production in simplified SUSY scenarios.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"EWP AIC background selection lacks an explicit spurious-signal systematic, so the exclusions may be sensitive to functional-form bias.","rationale":"The reader correctly identified the background functional form as the weakest assumption. My stress-test narrows this: the SP analysis uses discrete profiling, which covers the functional-form choice, but the EWP analysis uses AIC with a bias test that is not reported quantitatively and no explicit spurious-signal systematic is included. Since the top exclusions come from the EWP analysis, a nonzero spurious signal in the sensitive bins could shift the mass limits. This does not mean the result is wrong, but an accept should be conditional on a spurious-signal study demonstrating negligible bias, or alternatively on recomputing the EWP limits with the envelope method. Until that check is shown, the central exclusions rest on an unquantified assumption.","tokens_in":50040,"tokens_out":9417,"duration_ms":94302,"concrete_test":"Run a spurious-signal study for the 35 EWP search regions: generate pseudo-experiments from each candidate background functional form and from a diphoton+jets MC template, fit with the AIC-selected function, and record the fitted signal yield at 125 GeV; if the maximum fitted signal in the bins dominating the EWP expected limit exceeds roughly 20% of the expected background statistical uncertainty, the AIC selection is biased and the exclusion limits would need a spurious-signal systematic.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central exclusions for wino- and higgsino-like chargino-neutralino production are obtained with the EWP analysis (Section 8). The nonresonant background in the EWP analysis is modeled per bin with a functional form chosen by AIC after a bias test (Section 6). However, Section 7 reports systematic uncertainties only for the profiled floating background parameters; it does not assign an uncertainty for the discrete choice of functional form, such as the 'spurious signal' term used in the CMS H to gamma gamma measurement cited as Ref. [52]. If the AIC-selected form is biased low at m_gamma_gamma approximately 125 GeV in the bins that drive the limits (e.g., EWP 2, 9, 23), the observed counts in the 122-129 GeV window would be interpreted as a larger signal than is truly present, strengthening the exclusion limits. The bias test described in Section 6 only tests within the chosen functional family; it does not quantify the envelope of alternative shapes that the SP analysis's discrete profiling covers. Thus the EWP limits are load-bearing on an unquantified background-shape assumption.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents a search for supersymmetry in events with at least one Higgs boson decaying to two photons, using 77.5 fb^-1 of 13 TeV proton-proton collisions recorded by CMS. Two complementary analysis strategies are used: an electroweak-production-oriented analysis (EWP) with categories based on leptons, additional H/Z candidates, and razor variables, and a strong-production-oriented analysis (SP) with jet and b-jet counting plus mT2. The nonresonant diphoton and photon+jets background is modeled with functional fits, while the SM Higgs background is taken from simulation. No significant excess is found, and 95% CL exclusions are set on sbottom pair production, wino-like chargino-neutralino production, and higgsino-like GMSB production.","tokens_in":50289,"tokens_out":4162,"duration_ms":46157,"significance":"If the exclusions are correct, they extend previous CMS results by about 100 GeV for sbottom pair production and about 50 GeV for chargino-neutralino production, and they provide useful constraints on GMSB simplified models. The paper is commendably detailed: per-bin data yields, fitted backgrounds, signal expectations, and systematic uncertainties are shown in Tables 5-10, and the two-analysis strategy is a useful cross-check. The SP analysis uses discrete profiling for the background shape, which is a recognized way to cover functional-form uncertainty. The main weakness is that the EWP analysis does not explicitly assign a spurious-signal systematic to the AIC-selected background function, and the AIC bias test is not fully specified; since the EWP analysis drives the chargino-neutralino exclusions, this is a load-bearing point that needs to be addressed.","major_comments":[{"comment":"The EWP analysis selects the nonresonant background function using the AIC after a bias test, but no explicit spurious-signal uncertainty is assigned for the discrete choice of functional form. Section 7 propagates only the uncertainties in the profiled parameters of the selected function, even though the nonresonant background is stated to contribute 75-99% of the total uncertainty. Because the EWP analysis is used for the central wino-like and higgsino-like chargino-neutralino exclusions in Section 8, a low-side bias of the AIC-selected function near m_gamma_gamma = 125 GeV in sensitive bins (for example EWP 2, 9, and 23 in Table 7, where the observed yields are above the fitted background) would directly strengthen the reported exclusions. The authors should either add a spurious-signal systematic estimated from closure tests with injected signals, or demonstrate quantitatively that the AIC bias-test threshold bounds any such bias to a level negligible for the limits.","section":"Section 6 and Section 7, Table 7"},{"comment":"The description of the EWP bias test is incomplete: the manuscript does not specify the test statistic, the passing threshold, or the maximum allowed bias in the mass window around 125 GeV. The statement that the chosen functional form is 'adequate' is therefore not quantitatively supported. Since the central limits of the paper are derived from this background-modeling procedure, the authors should provide the bias-test details and, if possible, show the envelope of the bias across the 35 EWP search-region bins.","section":"Section 6"}],"minor_comments":[{"comment":"The first row for SP 25 is garbled: '53 252 53 662 ± 104 973 ± 68' should be formatted with consistent separators, e.g., observed 53,252, fitted background 53,662 ± 104, and SM Higgs background 973 ± 68.","section":"Table 6"},{"comment":"The caption reads 'two example search bin is shown' and should be corrected to 'two example search bins are shown'.","section":"Figure 2 caption"},{"comment":"The summary states the limits extend 'previous best CMS results [8,9]', but Ref. [9] is an ATLAS paper; the wording should distinguish the CMS and ATLAS comparisons or cite only Ref. [8] for CMS.","section":"Section 9"},{"comment":"The text describing the higgsino-like limit as 'chargino and neutralino (chi_1^0) masses of up to 290 and 230 GeV' is confusing because Figure 5 is plotted against the chi_1^0 mass; please clarify whether the quoted numbers refer to the neutralino mass or the chargino mass.","section":"Section 8 and Section 9"}],"recommendation":"major_revision","confidential_remarks":"This is a solid CMS search with detailed supporting tables, and the main result is likely correct. The missing spurious-signal systematic for the EWP AIC background selection is a standard internal-review point in CMS H->gamma gamma analyses and should be fixable with additional validation or a systematic term. I recommend major revision rather than rejection because the issue is localized to the uncertainty treatment of the background functional form and does not invalidate the analysis concept."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know three things about this one. First, it is a real extension, not a rehash: new lepton and H(bb)/Z(bb) categories, a separate strong-production mT2 strategy, and 77.5 fb-1 of 2016+2017 data push sbottom exclusions ~100 GeV and chargino-neutralino exclusions ~50 GeV past the earlier CMS razor paper. Second, the no-excess claim is supported by unusually transparent per-bin tables (data, fitted background, SM Higgs yield, uncertainties) for all 99 search regions. Third, the statistical machinery is mostly modern: discrete profiling for the SP analysis, unbinned fits, CLs with asymptotic formulas.\n\nWhat it does well: the categorization is thoughtful, the systematics are enumerated with sizes, and the paper is honest about which analysis gives the best limit in each model. The cross-section limits track NLO+NLL theory bands cleanly. For a null-result SUSY search this is competent, publishable work.\n\nNow the soft spots. The stress-test concern is real: the EWP analysis picks the background function with AIC after a bias test, but Section 7 only profiles the continuous parameters of that one chosen function. There is no spurious-signal term covering the discrete choice, unlike the SP analysis's envelope method or the H->gamma gamma measurement they cite. Since the wino and higgsino exclusions come from EWP bins, the central limits carry an unquantified functional-form bias. How bad? Probably modest: the diphoton window is narrow, the fits are done per bin, and the bias test catches gross failures within the family. But the correct fix is straightforward and I would want it before trusting the 235 and 290 GeV numbers to 1-2 GeV. A referee should ask for the spurious-signal study in the bins that drive those limits (EWP 2, 9, 23 look like the ones), or at least an envelope comparison for the EWP analysis.\n\nMinor issue: the exact family of background functions is described only vaguely (\"sums of exponentials, Bernstein polynomials...\") with no explicit list of the allowed variants and parameter counts. That made the AIC selection hard to reproduce from the text. Not fatal, but it is the kind of detail a serious referee would request.\n\nBottom line: the paper holds up on its central claim of no excess, and the exclusions are useful for SUSY model building. It deserves a serious referee, mostly to pin down that EWP background systematic. If you work in this area, cite it; the sbottom and wino mass reach numbers are now reference points.","headline":"A well-executed CMS SUSY search with genuinely new exclusions (sbottom ~530 GeV, wino ~235 GeV, higgsino ~290 GeV), but the EWP limits rest on a background functional-form choice that lacks an explicit spurious-signal systematic.","tokens_in":50770,"tokens_out":1270,"would_cite":true,"duration_ms":17987,"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":"A search for supersymmetry in Higgs-to-diphoton events at 13 TeV observes no excess and excludes sbottom masses below 530 GeV and chargino-neutralino masses below 235-290 GeV at 95% confidence.","keywords":["supersymmetry","Higgs boson to diphoton","bottom squark","chargino-neutralino","gauge-mediated supersymmetry breaking","razor variables","mT2","LHC"],"falsifier":"A concrete test would be to take a high-statistics diphoton control sample with negligible expected signal, split it into the same search region bins, inject a known artificial Higgs-like peak at 125 GeV, and check whether the background-family fit recovers the injected signal yield within the quoted uncertainty; a systematic bias larger than the quoted uncertainty would falsify the background-model assumption.","tokens_in":49860,"feed_emoji":"⚛️","tokens_out":10424,"duration_ms":94598,"temperature":0.7,"pith_summary":"The paper sets out to determine whether supersymmetric particles produced in pairs at the LHC leave a visible imprint when their decay chains pass through a Higgs boson that decays to two photons. Using $77.5\\,\\mathrm{fb^{-1}}$ of 13 TeV proton-proton collisions, the CMS experiment reconstructs Higgs candidates via the diphoton mass and classifies events into search regions sensitive to strong and electroweak SUSY production. The observed spectra agree with standard model backgrounds, and no statistically significant excess is found. Interpreting the null result in simplified models, the paper excludes bottom squark pair production for squark masses below 530 GeV with a lightest SUSY particle of 1 GeV, wino-like chargino-neutralino production for masses below 235 GeV, and higgsino-like chargino-neutralino production for neutralino masses below 290 GeV when the neutralino decays exclusively to a Higgs boson and a gravitino. If correct, these results extend the previous search by roughly 100 GeV (sbottom) and 50 GeV (chargino-neutralino).","feed_headline":"No SUSY signal in Higgs-to-diphoton decays, limits reach 530 GeV","feed_subtitle":"Bottom squarks below 530 GeV and charginos below 290 GeV are excluded at 95% confidence.","key_machinery":"The search's central tool is the Higgs-to-diphoton resonance tag: the diphoton invariant mass $m_{\\gamma\\gamma}$ is the discriminating observable, with SUSY signals producing a narrow peak near 125 GeV on top of a smoothly falling nonresonant background from standard model diphoton and photon+jets production. Events are divided into exclusive search regions using the number and flavor of leptons, b-tagged jet pairs compatible with $H\\to bb$ or $Z\\to bb$, and the kinematic variables $M_R$ and $R^2$ (razor variables, EWP analysis) or $m_{T2}$ and $p_T^{\\gamma\\gamma}/m_{\\gamma\\gamma}$ (SP analysis). The background shape is determined by fitting a family of analytic functions (sums of exponentials, Bernstein polynomials, Laurent series, and power laws) independently in each bin, selected by the Akaike information criterion for the EWP analysis or treated as a discrete nuisance via the envelope method for the SP analysis; signal and SM Higgs shapes are fixed from simulation using double Crystal Ball functions.","core_discovery":"The central claim is that no supersymmetry signal appears in the $\\mathrm{H}\\to\\gamma\\gamma$ final state at 13 TeV. After a simultaneous unbinned maximum-likelihood fit to the diphoton mass in all search regions, the data are consistent with the standard model prediction. The paper therefore reports exclusion limits at 95% confidence level for the simplified SUSY scenarios studied: bottom squark pair production with masses below 530 GeV (for a 1 GeV lightest SUSY particle), wino-like chargino-neutralino production in gauge-mediated SUSY breaking with chargino and neutralino masses below 235 GeV (1 GeV gravitino), and higgsino-like chargino-neutralino production in GMSB with neutralino masses below 290 GeV when $\\tilde\\chi^0_1\\to H\\tilde G$ is 100%, or below 230 GeV when $H\\tilde G$ and $Z\\tilde G$ are each 50%.","pith_inferences":["One consequence the paper leaves implicit: applying the same event categorization and background-fitting procedure to the full LHC Run 2 dataset (roughly twice the integrated luminosity) would likely extend the sbottom mass exclusion toward 600 GeV, provided the smooth background family remains adequate.","An independent cross-check would be to estimate the diphoton background with a data-driven control region (for example, $Z\\to e^+e^-$ events with the electrons treated as photons) instead of the analytic family; agreement between the two methods would strengthen the exclusion, while disagreement would expose a bias in the background model.","The categorization by leptons, $H\\to bb$, and $Z\\to bb$ tags is directly reusable for other new-physics searches that use a Higgs boson as a tag, such as top-squark pair production decaying through a Higgs boson, where similar final-state signatures appear."],"forward_implications":["If the result is correct, bottom squark pair production is excluded at 95% confidence for squark masses below 530 GeV when the lightest SUSY particle has mass 1 GeV.","Wino-like chargino-neutralino production in gauge-mediated SUSY breaking is excluded for chargino and neutralino masses below 235 GeV with a 1 GeV gravitino.","Higgsino-like chargino-neutralino production in GMSB is excluded for neutralino masses below 290 GeV when the $\\tilde\\chi^0_1\\to H\\tilde G$ branching fraction is 100%, and below 230 GeV when the $H\\tilde G$ and $Z\\tilde G$ branching fractions are each 50%.","The two-pronged analysis strategy (razor variables for electroweak production, $m_{T2}$ and jet/b-tag counting for strong production) extends the previous CMS result by about 100 GeV for sbottom and 50 GeV for chargino-neutralino mass reach."],"supporting_citations":[{"why":"The previous CMS search using razor variables; this analysis builds its electroweak event categories on that strategy and compares its improved sensitivity.","marker":"[8]"},{"why":"Introduces experimental signatures of low-energy gauge-mediated SUSY breaking, providing the $\\tilde\\chi^0_1\\to H\\tilde G$ decay scenario used in the models.","marker":"[4]"},{"why":"Analyzes Higgs and Z boson signatures of supersymmetry, motivating the H and Z decay modes of the neutralino in the GMSB simplified models.","marker":"[5]"},{"why":"The Akaike information criterion selects the background functional form in the EWP analysis.","marker":"[50]"},{"why":"The discrete profiling envelope method treats the background function as a discrete nuisance in the SP analysis.","marker":"[51]"},{"why":"A past CMS diphoton Higgs measurement that validates the accuracy of the two background-modeling approaches.","marker":"[52]"},{"why":"The CLs confidence-level computation used to derive the 95% exclusion limits.","marker":"[57]"},{"why":"Asymptotic formulae for likelihood-based tests provide the expected and observed limit evaluation.","marker":"[60]"},{"why":"Provides the 13 TeV NLO+NLL squark and gluino production cross sections used to interpret the sbottom exclusion.","marker":"[29]"},{"why":"Provides the chargino and neutralino production cross sections used for the electroweak interpretation.","marker":"[30]"}],"fun_headline_variants":["No SUSY signal in H→γγ decays, limits on squarks at 530 GeV","CMS excludes bottom squarks below 530 GeV in diphoton search","Search for SUSY in Higgs diphotons yields null result","Limits set on charginos and squarks from CMS diphoton data","Higgs to diphoton: no evidence for supersymmetry"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The result rests on the assumption that the standard model diphoton and photon-plus-jets background is accurately described by the chosen family of smooth analytic functions in the 125 GeV mass window, so that a genuine resonance is neither absorbed into the fit nor mimicked by it.","fun_headline_variants_meta":{"raw":{"variants":["No SUSY signal in H→γγ decays, limits on squarks at 530 GeV","CMS excludes bottom squarks below 530 GeV in diphoton search","Search for SUSY in Higgs diphotons yields null result","Limits set on charginos and squarks from CMS diphoton data","Higgs to diphoton: no evidence for supersymmetry"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000404,"raw_usage":{"total_tokens":2136,"prompt_tokens":1007,"completion_tokens":1129,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":623,"completion_tokens_details":{"reasoning_tokens":1030}},"tokens_in":623,"tokens_out":1129,"duration_ms":11678,"temperature":1.0,"reasoning_tokens":1030,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:38:25.068977+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete test would be to take a high-statistics diphoton control sample with negligible expected signal, split it into the same search region bins, inject a known artificial Higgs-like peak at 125 GeV, and check whether the background-family fit recovers the injected signal yield within the quoted uncertainty; a systematic bias larger than the quoted uncertainty would falsify the background-model assumption.","supporting_citations":[{"cited_title":"Experimental Signatures of Low Energy Gauge Mediated Supersymmetry Breaking","cited_arxiv_id":"hep-ph/9601367","evidence_quote":"Introduces experimental signatures of low-energy gauge-mediated SUSY breaking, providing the $\\tilde\\chi^0_1\\to H\\tilde G$ decay scenario used in the models."},{"cited_title":"Higgs and Z-boson Signatures of Supersymmetry","cited_arxiv_id":"hep-ph/9908482","evidence_quote":"Analyzes Higgs and Z boson signatures of supersymmetry, motivating the H and Z decay modes of the neutralino in the GMSB simplified models."}],"review_version":1}