{"id":"d2e51b2d-34c8-4839-99eb-6e1853a1e90b","arxiv_id":"1908.11741","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Existing LHC dilepton and four-lepton searches, reinterpreted for the minimal U(1)_{B-L} model, exclude gauge couplings down to about 5x10^-6 for a 0.25 GeV Z-prime at maximal Higgs mixing.","lead":"Researchers reinterpreted existing ATLAS and CMS muon searches to constrain the light Z-prime boson in the minimal U(1)_{B-L} model. They find that the four-lepton final state is sensitive to Z-prime masses as low as 0.25 GeV and can probe long-lived decays at the LHC.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"At the 0.25 GeV endpoint the 5e-6 g_B-L limit rests on an unvalidated efficiency transfer from CMS h to 4mu; the paper verifies alpha_gen only at 1 GeV and gives no public validation for the low-mass or displaced regions.","rationale":"The central construction is otherwise sound: the constraints are derived from published CMS and ATLAS limits rather than from a circular argument, the Higgs-mixing benchmark sin alpha = 0.3 is clearly stated and can be rescaled, and the branching-ratio treatment below 1 GeV is addressed with Darkcast. The fragile step is the transfer of the CMS h to 4mu detector efficiency to the B-L signal over the full mass range used for the headline claim. The reader identified this same efficiency transfer as the weakest assumption, but the most acute version of the problem is at the 0.25 GeV endpoint, where only a 1 GeV validation is reported, and in the 3.55 to 8.5 GeV extension, where no CMS limit table is public. The displacement concern is milder for the current 5e-6 bound because the corresponding lab-frame decay length is only about a centimeter, although the reconstruction efficiency for displaced tracks is still not documented. The proposed detector-level closure test would settle whether the efficiency transfer holds; until then a CONDITIONAL verdict is appropriate, with the current-data bounds plausible but not fully validated. No change to the reader's verdict is needed.","tokens_in":20452,"tokens_out":26166,"duration_ms":253601,"concrete_test":"Generate 10^4 pp to h to Z' Z' to 4mu events with the B-L UFO at m_Z' = 0.25, 1, 3.55, and 8.5 GeV for g_B-L at the claimed exclusion boundary, apply the CMS acceptance and displacement cuts (4.1)-(4.2) to compute alpha_gen, then pass the same samples through a CMS-like fast detector simulation (Delphes with muon trigger and reconstruction) and compute epsilon_total. Test whether epsilon_total / alpha_gen equals 0.60 within the CMS-quoted uncertainty at each mass, including displaced samples with average lab-frame decay length near 1 and 10 cm. If the ratio is not 0.60 at 0.25 GeV, rescale the quoted limit by sqrt(0.60 / epsilon_total) and re-evaluate the order-of-magnitude improvement over previous bounds.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The low-mass headline bound, g_B-L roughly 5e-6 at m_Z' = 0.25 GeV, is obtained by applying the CMS h to 4mu 'model-independent' detector efficiency, epsilon_det approximately 60 percent, to the B-L Z' signal. In Section 4 the only validation reported is a single sample: 'we have verified the reported alpha_gen by producing a sample of SM Higgs decaying to light Z' of 1 GeV and applying the cuts'. No validation is shown at m_Z' = 0.25 GeV, which is the point of the central claim, nor for the extension from 3.55 to 8.5 GeV where the published CMS limits stop, nor for displaced decays. The signal topology changes in exactly those corners: at 0.25 GeV the Z' is ultra-boosted and the muons are highly collimated, and at small g_B-L the decay vertices can be displaced; the 60 percent factor was demonstrated by CMS for its own benchmark samples and is not automatically universal. Because the limit scales as sqrt(1/epsilon), a true efficiency of 30 percent instead of 60 percent moves the 0.25 GeV bound from 5e-6 to about 7e-6, and a larger deficit erodes the claimed order-of-magnitude improvement. The extension above 3.55 GeV is a reimplementation using the cuts (4.1)-(4.2) and a single-bin background estimate rather than a published CMS limit table; it should not be relied on for the low-mass claim without public validation.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper investigates the reach of existing ATLAS and CMS searches for a light Z' in the minimal U(1)_{B-L} model, concentrating on Z' pair-production via the SM Higgs (pp→h→Z'Z'→4l), Drell-Yan Z' production, and final-state radiation from Z→4μ. The authors recast the CMS h→4μ search [46], the ATLAS h→4l search [47], the CMS L_μ−L_τ FSR search [48], and the CMS dimuon scouting search [45] into the (m_Z', g_B-L) parameter space for a fixed Higgs mixing sinα = 0.3. The central claim is that the four-lepton final state is sensitive to m_Z' as low as 0.25 GeV, with the CMS h→4μ analysis excluding g_B-L down to about 5×10^-6 at that mass, an order-of-magnitude improvement in the 0.25–1 GeV region, and that the same search can probe displaced Z' decays with lab-frame lengths up to about 10 cm.","tokens_in":20756,"tokens_out":7562,"duration_ms":63094,"significance":"If the central claim holds, the paper provides a valuable recast that helps fill a gap in the low-mass Z' parameter space and demonstrates the utility of Higgs-mediated production for probing light gauge bosons. The analytical formulas in Eqs. (2.10), (3.1), (3.2), (4.3), and (4.4) are clearly stated, and the use of public experimental limits with explicit scenario choices (m_N = m_Z'/3, sinα = 0.3) is a strength. However, the headline sensitivity at m_Z' = 0.25 GeV relies on an unvalidated transfer of the CMS detector efficiency to a kinematically extreme region, which is the main risk to the result. The extension of the CMS mass reach beyond 3.55 GeV also lacks public validation.","major_comments":[{"comment":"The reported validation of the generator-level acceptance α_gen is limited to a single sample at m_Z' = 1 GeV. The headline constraint g_B-L ≈ 5×10^-6 at m_Z' = 0.25 GeV (Sec. 5 and Fig. 8) is obtained by applying the CMS detector efficiency ε_detector ≈ 60% to a signal at that mass, where the Z' is ultra-boosted (βγ ~ 100) and the muons are highly collimated, and for small couplings the decays can be displaced with lab-frame length up to about 1 m. Because the limit scales as sqrt(1/ε), an efficiency of 30% instead of 60% would shift the 0.25 GeV bound from 5×10^-6 to about 7×10^-6, eroding the claimed order-of-magnitude improvement. This efficiency transfer is load-bearing and should be validated at the low-mass endpoint or treated with a conservative efficiency uncertainty.","section":"Sec. 4, CMS h→4μ paragraph"},{"comment":"The extension of the CMS h→4μ limits from the published mass range (0.25–3.55 GeV) to 8.5 GeV is a reimplementation using the cuts in Eqs. (4.1)–(4.2) and a single-bin background estimate of 9.90 ± 1.24_stat ± 1.84_syst events, rather than the model-independent upper limits reported by CMS. No public validation table is provided for this extension, so the limits displayed in Figs. 8 and 9 for m_Z' > 3.55 GeV rest on this unvalidated reimplementation. This is load-bearing for the claimed mass reach of the CMS h→4μ channel.","section":"Sec. 4, CMS h→4μ paragraph"},{"comment":"The statement that 'it is perfectly safe to use the analysis in this region' for the CMS h→4μ search in the displaced regime (m_Z' ≲ 0.5 GeV, g_B-L ≲ few×10^-6) is not supported by any validation of the efficiency transfer for decays with lab-frame displacement approaching the Lxy < 9.8 cm and Lz < 46.5 cm cuts. The CMS ε_detector ≈ 60% factor was demonstrated by the collaboration for its own benchmark samples and is not automatically universal for a B-L Z' with a different boost and decay-vertex distribution. This is particularly relevant for the HL-LHC projection, where the displaced region contributes to the sensitivity, and should be quantified before the long-lived sensitivity claim is made.","section":"Sec. 5, discussion of Fig. 8"}],"minor_comments":[{"comment":"The sentence 'The constraints from the ATLAS search [40] are somewhat stronger than the CMS search [58]' appears to refer to the wrong references; the h→4l searches are [47] and [46], while [40] and [58] are high-mass dilepton analyses. Please correct the citations.","section":"Sec. 6"},{"comment":"The phrase 'this analysis in not included' should read 'this analysis is not included'.","section":"Sec. 5"},{"comment":"The word 'Intheseesawlimit' is missing spaces and should be 'In the seesaw limit'.","section":"Sec. 2.1"},{"comment":"The sentence 'We have however used the model-independent limits given in Ref. [46]' is slightly confusing because the preceding sentence describes a χ^2 background estimate; consider clarifying that the χ^2 estimate is only used for the HL-LHC projection and not for the current limits.","section":"Sec. 4, CMS h→4μ paragraph"}],"recommendation":"major_revision","confidential_remarks":"The paper is a straightforward and useful recast of existing searches, but the unvalidated efficiency transfer at the low-mass endpoint is a genuine concern that should be addressed before publication. The reference mismatch in Sec. 6 is a simple fix. I recommend major revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Punchline: this is a competent recasting paper that gives the first combined LHC interpretation of Higgs-mediated Z' pair production, FSR, and dilepton searches in the minimal B-L model. The headline exclusion of g_B-L ~ 5e-6 at m_Z' = 0.25 GeV is plausible, but it rests on an efficiency transfer that the authors only validate at 1 GeV. I would send this to review with a request for public validation.\n\nWhat's new: the individual ingredients are known, but the combination fills a previously under-constrained window between 0.25 and 1 GeV and explicitly maps out the displaced-decay region. The analytical cross sections and width formulas are clearly stated, and the limit conversions are coherent. There is no circularity: they compare independent theory predictions to external upper limits, with fixed scenario inputs (sin alpha = 0.3, m_N = m_Z'/3) stated as such. The heavy-neutrino discussion in Sec. 6 is a useful consequence, showing the new constraints suppress the previously claimed discovery potential.\n\nSoft spots: the main one is the efficiency transfer for the CMS h->4mu analysis. The paper verifies alpha_gen only for a 1 GeV Z' and then applies a 60% detector efficiency down to 0.25 GeV and out to displaced decays. The stress-test note is right that this is not validated; a factor-two error in efficiency shifts the 0.25 GeV bound from 5e-6 to 7e-6, which does not kill the order-of-magnitude improvement but should be documented. The extension above 3.55 GeV is a reimplementation using a single-bin background estimate, not a published CMS limit table; that is a separate and less central issue. The HL-LHC projections are simple luminosity rescaling with no systematics, so they should be labelled as optimistic. No public code or processed limit tables are provided, which makes independent checking harder. These are all addressable rather than fatal.\n\nOverall: a useful phenomenology paper, not a paradigm change. The citation pattern is appropriate, including the self-citations to the authors' own B-L and heavy-neutrino work, which are directly relevant. I would bring it to the reading group, and I would probably cite it in B-L related work. For peer review: accept with requests for validation of the efficiency transfer, a limit table for the reimplemented region, and a note on systematics in the projections.","headline":"Solid recasting paper with a plausible low-mass B-L exclusion, but the 0.25 GeV headline limit leans on an efficiency transfer validated only at 1 GeV; deserves review with a request for public validation.","tokens_in":21366,"tokens_out":3891,"would_cite":true,"duration_ms":37683,"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":"Four-lepton Higgs decays at the LHC can reveal a B-L Z' boson as light as 0.25 GeV.","keywords":["B-L gauge boson","Z' boson","Higgs portal","four-lepton final state","LHC recasting","long-lived particle","displaced vertex","seesaw mechanism"],"falsifier":"Recalculate the CMS $h\\to 4\\mu$ selection with a full detector simulation for $pp\\to h\\to Z'Z'\\to 4\\mu$ with $m_{Z'}=0.25$ GeV, $g_{B-L}=10^{-5}$, and $\\sin\\alpha=0.3$, and compare the accepted signal yield with the efficiency-transfer prediction; if the yield is smaller than predicted by more than the quoted uncertainty, the $5\\times 10^{-6}$ bound shifts. A simpler check would be a public table of generator-level acceptance times detector efficiency for the mass grid, since the paper gives none for the extension above 3.55 GeV.","tokens_in":20205,"feed_emoji":"⚛️","tokens_out":17528,"duration_ms":141096,"temperature":0.7,"pith_summary":"The paper argues that the minimal $U(1)_{B-L}$ extension of the Standard Model — a simple gauge route to neutrino masses — can be probed with existing LHC searches much more directly than has been assumed: the CMS search for Higgs decays to four muons is sensitive to a $Z'$ as light as 0.25 GeV. With the currently allowed Higgs mixing $\\sin\\alpha = 0.3$, that search excludes $g_{B-L}$ down to about $5\\times 10^{-6}$, closing much of the previously unconstrained window below 1 GeV and improving existing limits there by roughly an order of magnitude. The paper also shows that in this region the $Z'$ is often long-lived in the lab frame, with mean decay lengths up to about 10 cm, so the same four-lepton searches double as displaced-vertex searches. Combined with ATLAS four-lepton, CMS final-state-radiation, and CMS data-scouting dilepton analyses, the result maps the $m_{Z'}$--$g_{B-L}$ plane from 0.25 GeV to tens of TeV.","feed_headline":"Four-muon Higgs decays can catch Z' bosons down to 0.25 GeV","feed_subtitle":"Existing LHC four-lepton data already exclude B-L gauge couplings down to about 5e-6.","key_machinery":"The engine of the argument is the minimal $U(1)_{B-L}$ model: a Standard Model extension with one extra $U(1)$ gauge group under which quarks and leptons carry baryon-minus-lepton number, a singlet scalar $\\chi$ whose vacuum expectation value gives the $Z'$ its mass $m_{Z'} = 2g_{B-L}\\langle\\chi\\rangle$, and three right-handed neutrinos for anomaly cancellation and seesaw neutrino masses. The production mechanism that carries the result is the Higgs portal: mixing between the SM Higgs and $\\chi$, parametrized by $\\sin\\alpha$, allows $pp\\to h\\to Z'Z'$ with a partial width proportional to $(g_{B-L}\\sin\\alpha/m_{Z'})^2$. The recasting is carried by the CMS $h\\to 4\\mu$ analysis's reported signal-model-independent detector efficiency of about 60%, which the authors use to extend the quoted 0.25--3.55 GeV limits up to 8.5 GeV, together with generator-level acceptance cuts on muon transverse momentum, pseudorapidity, and displacement $L_{xy}<9.8$ cm, $L_z<46.5$ cm.","core_discovery":"The central discovery is that Higgs-mediated production $pp\\to h\\to Z'Z'\\to 4\\mu$, not direct Drell-Yan production, is the channel that opens up light $Z'$ bosons at the LHC. Recasting the CMS $h\\to 4\\mu$ search with its model-independent limits, the authors find that the four-lepton final state is sensitive to $m_{Z'}$ as low as 0.25 GeV and, at $\\sin\\alpha = 0.3$, constrains $g_{B-L}$ to about $5\\times 10^{-6}$; the limit weakens to $1.8\\times 10^{-4}$ at $m_{Z'} = 8.5$ GeV. Because the $Z'$ is boosted in Higgs decays, it can be displaced in the lab even when its proper lifetime is short, and the CMS search's 9.8 cm transverse displacement cut lets the bound cover mean lab-frame decay lengths up to about 10 cm. The ATLAS four-lepton search gives stronger limits where it applies, but only for prompt decays and with gaps from QCD resonances; the CMS dimuon scouting search takes over above about 50 GeV, while the final-state-radiation channel is the weakest.","pith_inferences":["If the efficiency transfer holds, the same CMS $h\\to 4\\mu$ recasting should apply to any Higgs-coupled light gauge boson with a muonic decay, making the 0.25 GeV frontier a generic test of Higgs portals rather than a $B-L$-specific result.","The analysis assumes zero kinetic mixing between $U(1)_{B-L}$ and hypercharge; opening that mixing would change both production and decay of the $Z'$ and would likely alter the low-mass limits, so the $5\\times 10^{-6}$ bound should be read as specific to the minimal model.","A dedicated search with a transverse displacement cut larger than the 9.8 cm used by CMS could push sensitivity below $g_{B-L}\\sim 10^{-6}$ for $m_{Z'}<0.5$ GeV, where the paper shows the $Z'$ lifetime grows while the Higgs-mediated production cross section remains non-negligible.","A future precise measurement of the Higgs mixing angle, for example at a lepton collider, would break the $\\sin\\alpha$--$g_{B-L}$ degeneracy and turn this combined search into a direct measurement of the gauge coupling."],"forward_implications":["Existing LHC Run 2 data, not future searches, already exclude a $B-L$ $Z'$ with $m_{Z'}$ around 0.25 GeV and coupling near $5\\times 10^{-6}$, a region previously considered hard to reach.","At the High-Luminosity LHC, the same four-lepton searches will push the coupling limits lower across 0.25--8.5 GeV, with the gain muted in the displaced region below roughly 0.5 GeV.","The improved bounds nearly close the window for observable heavy-neutrino production through the $Z'$ or Higgs in this model: $pp\\to Z'\\to NN$ cross sections fall to at most a few femtobarns, and Higgs-mediated heavy-neutrino rates are suppressed by about three orders of magnitude relative to previously considered values.","Because the Higgs-portal limits scale with $g_{B-L}\\sin\\alpha$, the result can be rescaled to any assumed mixing: for example, at $m_{Z'}=1$ GeV the coupling limit moves from $3\\times 10^{-5}$ at $\\sin\\alpha=0.3$ to $5\\times 10^{-5}$ at $\\sin\\alpha=0.2$."],"supporting_citations":[{"why":"Supplies the model-independent upper limits, the acceptance cuts, and the reported approximately model-independent 60% detector efficiency that the paper extends up to 8.5 GeV.","marker":"[46]"},{"why":"Supplies the prompt four-lepton signal-strength limits that set the strongest constraints in the 1--60 GeV mass range, with gaps from QCD resonances.","marker":"[47]"},{"why":"Supplies the CMS Z->4mu final-state-radiation limits, converted to a B-L coupling bound via the relation between the L_mu-L_tau and B-L gauge couplings.","marker":"[48]"},{"why":"Supplies the CMS dimuon data-scouting resonance limits that become the best constraints above about 50 GeV.","marker":"[45]"},{"why":"Defines the minimal $U(1)_{B-L}$ model with the $Z'$, the singlet scalar, and the right-handed neutrinos that the whole calculation assumes.","marker":"[3]"},{"why":"Provides the current constraints on the Higgs mixing angle that justify taking $\\sin\\alpha=0.3$ as the maximally allowed value.","marker":"[19]"},{"why":"Supplies the low-mass branching-ratio treatment and the existing beam-dump and LHCb limits that define the previously unconstrained region the paper improves.","marker":"[16]"},{"why":"Supplies the NLO model implementation used for the event generation of Higgs-mediated Z' pair production.","marker":"[51]"}],"fun_headline_variants":["Higgs decays expose light Z' down to 0.25 GeV","Four-muon Higgs channel bounds Z' coupling at 5e-6","LHC Higgs portal opens new window to long-lived Z'","Searching for light Z' via Higgs-mediated 4mu at LHC","Higgs-produced Z' pairs: a new LHC probe of B-L"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole low-mass bound rests on the assumption that the detector efficiency reported by the CMS $h\\to 4\\mu$ analysis, found to be roughly signal-model independent, transfers unchanged to $B-L$ $Z'$ events at every mass used in the paper, including $m_{Z'}=0.25$ GeV and decays with muons displaced up to 9.8 cm.","fun_headline_variants_meta":{"raw":{"variants":["Higgs decays expose light Z' down to 0.25 GeV","Four-muon Higgs channel bounds Z' coupling at 5e-6","LHC Higgs portal opens new window to long-lived Z'","Searching for light Z' via Higgs-mediated 4mu at LHC","Higgs-produced Z' pairs: a new LHC probe of B-L"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000325,"raw_usage":{"total_tokens":1865,"prompt_tokens":1033,"completion_tokens":832,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":649,"completion_tokens_details":{"reasoning_tokens":734}},"tokens_in":649,"tokens_out":832,"duration_ms":7912,"temperature":1.0,"reasoning_tokens":734,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T10:08:42.464241+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recalculate the CMS $h\\to 4\\mu$ selection with a full detector simulation for $pp\\to h\\to Z'Z'\\to 4\\mu$ with $m_{Z'}=0.25$ GeV, $g_{B-L}=10^{-5}$, and $\\sin\\alpha=0.3$, and compare the accepted signal yield with the efficiency-transfer prediction; if the yield is smaller than predicted by more than the quoted uncertainty, the $5\\times 10^{-6}$ bound shifts. A simpler check would be a public table of generator-level acceptance times detector efficiency for the mass grid, since the paper gives none for the extension above 3.55 GeV.","supporting_citations":[{"cited_title":"collaboration,Search for a narrow resonance decaying to a pair of muons in proton-proton collisions at 13 TeV, Tech","cited_arxiv_id":null,"evidence_quote":"Supplies the CMS dimuon data-scouting resonance limits that become the best constraints above about 50 GeV."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the minimal $U(1)_{B-L}$ model with the $Z'$, the singlet scalar, and the right-handed neutrinos that the whole calculation assumes."}],"review_version":1}