{"id":"181a7c20-1652-4917-afac-534fb16d03af","arxiv_id":"2412.11438","paper_version":1,"verdict":"REJECT","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"high","formal_verification":"none","parameter_count":5,"one_line_summary":"Only a fine-tuned dark Z model with cancelled electron couplings survives the combined constraints, but its muon g-2 contribution is orders of magnitude too large.","lead":"This paper tests a dark photon/dark Z model against B and K meson decay data, finds it largely excluded, and argues that only a fine-tuned variant with suppressed electron couplings survives. The result is a set of constraints for model builders, but the surviving scenario is undermined by the paper's own muon g-2 calculation and an unexplained inconsistency with the K to muon plus invisible bound.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Case C's fine-tuned electron cancellation cannot work: the vector-only ge_D of Eq. (6) cannot cancel the axial electron coupling induced by εZ, so APV constraints remain and the central claim fails.","rationale":"The reader's verdict of REJECT is supported, but the strongest reason is even more direct than the reader's stated weakest assumption. The reader noted that the fine-tuning lacks a mechanism and would fail under mistuning. I find that even exact tuning of a vector-only ge_D cannot cancel the axial electron coupling induced by the Z mass mixing. This is a model-level inconsistency, not a parameter tuning issue: the cancellation required for Case C to bypass APV is impossible with the interaction of Eq. (6). The paper's Eqs. (2)–(3) make the axial coupling explicit in C10, and Section 4.7's APV bounds constrain εZ. Therefore the central claim that Case C survives APV is not just fragile; it is internally contradicted. A secondary, also serious, contradiction is the gμ_D = 0.033 best fit versus the gμ_D < 0.01 bound in Section 4.5 for K+→μ+ν ZD with invisible decay, which the paper does not reconcile for Case C. Both issues are checkable from the text, so confidence in rejection is high. The paper does use established tools (Peng4BSM@LO, flavio) and provides a useful survey of constraints on the simpler cases, but the flagship conclusion fails. No independent code or data is shipped, making independent verification of the numerical fits impossible; however, the internal inconsistencies suffice.","tokens_in":13506,"tokens_out":11115,"duration_ms":90156,"concrete_test":"Compute the effective ZD-electron couplings from Eqs. (2)–(3) using the Case C Lagrangian of Eq. (6): fix ge_D by requiring the vector coefficient V_e = eε + (g/cosθW) εZ g_V^e + ge_D = 0, then evaluate A_e = (g/cosθW) εZ g_A^e. For any εZ > 0, A_e is nonzero. Next, apply the APV constraint of Section 4.7 to this A_e by repeating the weak-charge calculation without the assumed axial cancellation, scanning the Case C 2σ region (MZD ≈ 30 MeV, gμ_D ≈ 0.033, ε and εZ as used in Fig. 10). If APV excludes the region, the central claim fails; this test directly checks whether the proposed fine-tuning can actually remove the electron coupling.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—that only the fine-tuned Case C remains consistent with all experimental constraints—rests on the assumption in Section 3 that ge_D can be chosen to cancel the mixing-induced ZD coupling to electrons, making all electron-mode observables SM-only. This assumption is internally inconsistent with the model's own couplings. The mass-mixing term of Section 2 gives ZD a coupling to the SM Z current, L ⊃ (g/cosθW) εZ Z_D^μ J_μ^Z, so the mixing-induced electron coupling contains both vector and axial pieces: V_e = eε + (g/cosθW) εZ g_V^e and A_e = (g/cosθW) εZ g_A^e, with g_A^e = -1/2. The direct coupling added in Eq. (6), ge_D \\bar e γ^α e Z_α^D, is purely vector and can only shift V_e; it cannot change A_e. Therefore, whenever εZ ≠ 0, the axial electron coupling remains, electron observables are not SM-only, and the APV constraints of Section 4.7 remain active. Since εZ is required to be nonzero for the muon-mode fit (e.g., for Bs→μ+μ− via C10), Case C cannot evade APV by this mechanism. The paper's statements in Sections 4.7 and 5 that Case C is unconstrained by APV are thus not supported by its own Lagrangian. Independently of this, the best-fit gμ_D = 0.033 appears to violate the authors' own gμ_D < 0.01 bound from K+→μ+ν ZD with invisible decay (Section 4.5), and no recast for Case C is given. Either issue alone invalidates the abstract's claim.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies a dark U(1)_D model with a dark photon/dark Z boson in the mass range 10 MeV to 2 GeV, applied to flavor-changing neutral-current decays of B and K mesons. Three cases are considered: the base model (Case A), an extension with a direct vector coupling to muons (Case B), and an extension with direct couplings to both muons and electrons, where the electron coupling is fine-tuned to cancel the mixing-induced coupling to electrons (Case C). The authors fit b→sℓ+ℓ− data using the packages Peng4BSM@LO and flavio, and confront the resulting parameter space with constraints from B_s mixing, B_s→μ+μ−, B→K(*)νν̄, kaon decays, atomic parity violation, neutrino trident, and other searches. The central claim, stated in the abstract, is that only Case C remains consistent with all experimental constraints, with a best fit at M_ZD = 30.2 MeV and gμ_D = 0.033.","tokens_in":13863,"tokens_out":5146,"duration_ms":45500,"significance":"If the central claim were correct, the paper would identify a narrow window for a light dark Z with fine-tuned electron coupling that evades low-energy constraints while improving the fit to b→sμ+μ− data. The paper has the merit of combining many constraints—APV, kaon decays, B_s mixing, trident, W width, and LHCb dark-photon searches—and of including hadronic Z_D decays through vector-meson dominance. However, the central claim is undermined by internal inconsistencies: the proposed fine-tuned cancellation cannot remove the axial electron coupling induced by mass mixing, the best-fit gμ_D appears to violate the paper's own K+→μ+ν+invisible bound, and the predicted muon g−2 exceeds the measured discrepancy by orders of magnitude. These are load-bearing flaws, not presentation issues, and they invalidate the abstract's conclusion.","major_comments":[{"comment":"The fine-tuned cancellation of the Z_D coupling to electrons in Case C is impossible with the interaction written in Eq. (6). The mass-mixing term in Section 2 gives Z_D a coupling to the SM Z current, L ⊃ (g/cosθ_W) ε_Z Z_D^μ J_μ^Z, which contains both vector and axial electron pieces, with axial coefficient (g/cosθ_W) ε_Z g_A^e and g_A^e = −1/2. The direct coupling added in Eq. (6), g_e^D \\bar e γ^α e Z_α^D, is purely vector and can only shift the vector coefficient; it cannot change the axial coupling. Therefore, for any ε_Z ≠ 0, an axial electron coupling remains, electron-mode observables are not SM-only, and the APV constraints of Section 4.7 remain active. The statements in Sections 4.7 and 5 that Case C is unconstrained by APV are not supported by the model's own Lagrangian.","section":"Section 3, Eq. (6); Section 4.7"},{"comment":"Section 4.5 states that for a directly coupled muonphilic Z_D, the K+ → μ+ν Z_D (→ invisible) branching fraction requires gμ_D < 0.01. In Case C, the electron coupling is cancelled, so for M_Z_D below the dimuon threshold the Z_D decays invisibly (to neutrinos or dark-sector particles), and the same argument should apply. However, the Case C best fit in Section 5 has gμ_D = 0.033, which exceeds this bound. No recast of the K+ → μ+ν+invisible limit for Case C is provided, so the claim in Section 5 that the Case C parameter space 'remains consistent with ... K → μ + invisible' is unsupported.","section":"Section 4.5 vs. Section 5"},{"comment":"The predicted contributions to the muon anomalous magnetic moment at the best-fit points are a_A^μ = −3.45×10^−7, a_B^μ = 7.7×10^−4, and a_C^μ = 7.38×10^−6. These exceed the measured discrepancy Δa_μ ≈ 251(59)×10^−11 by one to several orders of magnitude. The paper acknowledges this and appeals to an additional 'dark charged scalar' to cancel the contribution, but that particle is not part of the model and its parameters are not specified. As presented, the model itself is not consistent with all experimental constraints, contradicting the abstract's claim that Case C 'remains consistent with all experimental constraints.'","section":"Section 6"},{"comment":"Section 4.1 concludes that 'ε_Z ≥ 0.001 is disallowed for M_ZD ≤ 60 MeV' from B_s−B̄_s mixing, yet the Case A best-fit point in Section 5 is M_ZD = 10.07 MeV with ε_Z = 0.002, which lies in the disallowed region. This inconsistency is not addressed, and it casts doubt on how the fit results are combined with the low-energy constraints. If the disallowed statement is meant at a different confidence level or under different assumptions, the text must state this explicitly.","section":"Section 4.1 vs. Section 5"}],"minor_comments":[{"comment":"The effective Hamiltonian expressions contain undefined symbols such as V_c^ν, E0,c couplings, and M1,c couplings; the reader is not told how these are computed or normalized before they are used in Eqs. (2) and (3).","section":"Section 2"},{"comment":"The parameters ρ_d and κ_d are introduced but never defined; the formulas relating them to ε, ε_Z, and M_Z_D are omitted, which makes the APV constraint difficult to reproduce.","section":"Section 4.7"},{"comment":"The fit paragraph refers to 'a further extension with an axial-vector coupling of the dark Z to muons (Case C)', but Eq. (6) defines Case C with only vector couplings; this terminology inconsistency should be resolved.","section":"Section 5"},{"comment":"The phrase 'in the mass range 0 < M_X < 33.9 MeVperimental bound is used' appears to be a typographical error; it should read something like 'in the mass range 0 < M_X < 33.9 MeV; the experimental bound is used...'.","section":"Section 4.6"},{"comment":"Reference [42] is incomplete: 'Review of Particle Physics, .' lacks volume, year, and article number; also, the text has several other typos (e.g., 'flavour' vs. 'flavor' is acceptable, but 'anfluenced' appears broken) that should be corrected.","section":"References"}],"recommendation":"reject","confidential_remarks":"The two central technical flaws—the impossibility of the vector-only cancellation removing the axial electron coupling, and the apparent violation of the paper's own K+→μ+ν+invisible bound by the Case C best fit—are not local fixes; they invalidate the main conclusion. Even if the authors added an axial direct coupling to make the cancellation possible, the model would require an additional fine-tuning parameter and the APV constraints would not disappear as cleanly as claimed. The muon g−2 problem further indicates that the model, in isolation, is not consistent with all constraints. The paper would require a substantial reformulation and a re-analysis of the viable parameter space before it could be considered for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Thanks for the report. I read the paper, and I agree with the reject verdict, with one additional technical reason that I think is decisive.\n\nThe paper does some useful things. It extends the dark Z analysis of Xu [16] by including hadronic decays and the q^2-independent monopole operator, and it runs a flavio-based fit to b→sℓ+ℓ− and related modes in three benchmark scenarios. The plots and constraint compilation are clear, and the treatment of B_s mixing, B_s→μ+μ−, and APV is a reasonable survey. If the central claim were right, the Case C parameter point would be of interest to the dark sector community.\n\nBut the central claim isn't right. Three problems:\n\n1) The fine-tuned electron cancellation in Case C cannot work as stated. The mixing-induced coupling from the Lagrangian in Section 2 includes both vector and axial electron currents, proportional to (g/cosθW) εZ g_A^e. The direct coupling added in Eq. (6) is purely vector. So g_e^D can at best cancel the vector piece; the axial piece remains whenever εZ ≠ 0, which is required for the muon-mode fit. APV constraints therefore still apply to Case C. The paper's statement that Case C is unconstrained by APV is inconsistent with its own theory.\n\n2) The paper contradicts itself on K+→μ+ν+invisible. Section 4.5 gives g_mu^D < 0.01 when Z_D decays invisibly. In Case C, Z_D at 30 MeV has no electron coupling and is below the dimuon threshold, so it decays invisibly. The best fit g_mu^D = 0.033 is above that bound, and no recast for Case C is provided.\n\n3) The abstract's 'consistent with all experimental constraints' is false on the paper's own numbers. Section 6 reports a_C^μ = 7.38×10^{-6}, thousands of times the measured (g-2)_μ discrepancy. The paper acknowledges this in passing but does not fold it into the fit or summary.\n\nThese are load-bearing, not cosmetic. The underlying matrix-element calculation might still be usable, and the monopole correction is worth checking. But as submitted, the paper's headline conclusion is unsupported. It would need substantial revision—either a consistent mechanism for cancelling the axial electron coupling (a direct axial term or an alternative), or a redrawing of the conclusions to admit that Case C does not satisfy all constraints.\n\nI would not bring this to reading group, and I would not cite it in its current form. A serious referee could in principle salvage the technical part, but the internal contradictions are clear enough that I'd desk-reject with an invitation to resubmit after fixing the model.","headline":"The paper's central claim fails on its own Lagrangian: the vector-only electron coupling cannot cancel the axial term from mass mixing, and the best-fit gμ_D=0.033 violates the paper's own K+→μ+invisible bound.","tokens_in":14436,"tokens_out":5466,"would_cite":false,"duration_ms":45951,"reading_group":"maybe","serious_thinker":"no","would_accept_peer_review":false},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper claims that of three dark-Z scenarios, only one—a roughly 30 MeV mediator with a direct muon coupling and a fine-tuned cancellation of its electron coupling—survives every current constraint while improving the fit to b→s μ+μ−…","keywords":["dark photon","dark Z boson","flavor-changing neutral currents","B meson decays","K meson decays","lepton flavor universality","atomic parity violation","muon anomalous magnetic moment"],"falsifier":"A dedicated NA62 search for K+→μ+ν+invisible sensitive below the rate predicted at gμD = 0.033 and MZD = 30 MeV would falsify the surviving region; failing that, a cesium weak-charge measurement more precise than the tolerance of the tuned cancellation would bring the atomic parity violation exclusion back.","tokens_in":13195,"feed_emoji":"⚛️","tokens_out":7976,"duration_ms":76046,"temperature":0.7,"pith_summary":"This paper asks whether a light dark photon or dark Z boson produced in rare flavor-changing B and K decays can explain the observed discrepancies in b→s μ+μ− rates, and which parameter choices survive all other low-energy constraints. It finds that the basic mixing-only model improves the fit to the B-decay data but is excluded mainly by atomic parity violation, K+→μ+ν+invisible, and Bs−B̄s mixing. Adding a direct muon coupling improves the fit further, but the best-fit coupling is excluded by kaon radiative decays and the W-boson width. The only surviving option is a dark Z with mass near 30 MeV, direct muon coupling gμD around 0.033, and an electron coupling tuned to cancel the mixing-induced ZD-electron interaction, leaving all electron-mode observables Standard-Model-like. A sympathetic reader would take this as a map of where a light-mediator explanation of the B anomalies can still live.","feed_headline":"Fine-tuned dark Z survives all B and K decay constraints","feed_subtitle":"A 30 MeV mediator with muon coupling 0.033 fits the b→s μ+μ− data where simpler dark-photon models fail.","key_machinery":"The load-bearing object is the massive dark gauge boson ZD from a broken U(1)D dark sector, coupled to the Standard Model through kinetic mixing (parameter ε) and, in the dark-Z variant, mass mixing (parameter εZ). Flavor-changing neutral-current transitions b→sZD, s→dZD, and b→dZD are generated at one loop, and the resulting effective Hamiltonian feeds q2-dependent Wilson coefficients C9 and C10 through a ZD propagator that becomes resonant when the mediator is on shell. The decisive mechanism in the surviving scenario is a fine-tuned direct coupling geD chosen to cancel the mixing-induced ZD-e+e− vertex, so every electron-mode observable reduces to the Standard Model and atomic parity violation no longer applies; the muon coupling gμD then controls the B and K anomalies.","core_discovery":"On the paper's own terms, the central discovery is a parameter-space scan: of three versions of a U(1)D dark Z in the 10 MeV to 2 GeV range, only the version with a direct vector coupling to muons and a fine-tuned direct coupling to electrons that cancels the mixing-induced ZD-e+e− interaction is consistent with the full set of constraints. The fit to binned b→s μ+μ− data prefers MZD = 30.2 MeV and gμD = 0.033, a region that fits the data at 2σ and also evades atomic parity violation, K→μ+invisible, neutrino trident production, Bs mixing, and LHCb dark-photon searches. The muon-only extension needed gμD = 0.28 and was excluded by K+→μ+νX and the W width; the base model was excluded by atomic parity violation below roughly 30 MeV. The paper also notes that at its surviving point the dark Z overshoots the muon g−2 discrepancy, so a complete model needs an additional negative contribution.","pith_inferences":["The required cancellation is not protected by any symmetry in the paper; radiative corrections will generically regenerate a small ZD-electron coupling, so a UV-complete version must enforce the tuning and would predict tiny but nonzero electron couplings that improved APV or electron-beam searches could detect.","The same tuning trick could rescue other light-mediator explanations of the B anomalies that are currently excluded by atomic parity violation or electron-mode experiments, shifting the decisive tests to muon-mode channels such as LHCb dimuon searches and CCFR trident production.","The paper's own Section 4.5 quotes gμD < 0.01 from K+→μ+ν+invisible, while the surviving fit uses gμD = 0.033; a dedicated recast of that NA62 limit with the Case C invisible-width assumptions would decide whether the surviving region is as wide as the summary suggests.","If a dark Z with gμD around 0.033 exists, its positive contribution to (g−2)μ is large enough to require a second, negative contribution elsewhere, making the model testable through the correlation between a future B-anomaly measurement and a future muon g−2 measurement."],"forward_implications":["If the surviving scenario is correct, a light ZD with MZD near 30 MeV and gμD around 0.033 is compatible with all current B, K, APV, trident, and collider bounds, so future b→s μ+μ− and kaon-decay data will probe exactly this parameter region.","The muon-only extension is dead: its best-fit point gμD = 0.28 violates the K+→μ+νX and W-width constraints, so any future light-mediator explanation of the B anomalies must keep the muon coupling at the few-percent level and add another ingredient.","Electron-mode observables no longer discriminate among models once the electron coupling is tuned away, so the decisive experimental handles become muon-mode decays, rare kaon decays, and precision width measurements.","The surviving point leaves the muon g−2 discrepancy unresolved and in fact overshoots it, implying that additional new physics with a negative contribution must accompany the dark Z in any complete model."],"supporting_citations":[{"why":"introduces the kinetic-mixing term that defines the dark photon's coupling to the electromagnetic current.","marker":"[1]"},{"why":"builds the dark-Z mass-mixing scenario and its parity-violation implications that drive the APV constraint.","marker":"[3]"},{"why":"proposes the light-resonance interpretation of the low-q2 RK* anomaly that the fit is testing.","marker":"[9]"},{"why":"earlier dark-Z flavor calculation whose missing q2-independent monopole contribution this paper corrects.","marker":"[16]"},{"why":"supplies the one-loop penguin amplitudes for the b→s, s→d, and b→d FCNC transitions.","marker":"[20]"},{"why":"gives the experimental R ratios and PDG values used for ZD hadronic widths and Bs mixing.","marker":"[24]"},{"why":"provides the LHCb Bs→μ+μ− measurement that constrains εZ at the 3σ level.","marker":"[33]"},{"why":"sets the B+→K+νν upper bound used to check the on-shell ZD→invisible channel.","marker":"[37]"},{"why":"provides the differential B0→K*0 μ+μ− branching fractions used in the q2-bin fit.","marker":"[44]"}],"fun_headline_variants":["Dark Z survives only with fine-tuned electron coupling","Only fine-tuned dark Z passes B and K meson tests","Dark Z at 30 MeV survives with fine-tuned couplings","Dark Z with muon coupling 0.033 fits data, evades all limits"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The model survives only because the direct electron coupling geD is fine-tuned to cancel the mixing-induced ZD-electron interaction exactly, and the paper offers no symmetry or mechanism enforcing that cancellation; implicitly, it also assumes the K+→μ+ν+invisible bound of Section 4.5 (gμD < 0.01) does not apply to the fitted point gμD = 0.033.","fun_headline_variants_meta":{"raw":{"variants":["Dark Z survives only with fine-tuned electron coupling","Only fine-tuned dark Z passes B and K meson tests","Dark Z at 30 MeV survives with fine-tuned couplings","Dark Z with muon coupling 0.033 fits data, evades all limits"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000737,"raw_usage":{"total_tokens":3321,"prompt_tokens":1000,"completion_tokens":2321,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":616,"completion_tokens_details":{"reasoning_tokens":2246}},"tokens_in":616,"tokens_out":2321,"duration_ms":17013,"temperature":1.0,"reasoning_tokens":2246,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T14:56:30.609912+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A dedicated NA62 search for K+→μ+ν+invisible sensitive below the rate predicted at gμD = 0.033 and MZD = 30 MeV would falsify the surviving region; failing that, a cesium weak-charge measurement more precise than the tolerance of the tuned cancellation would bring the atomic parity violation exclusion back.","supporting_citations":[{"cited_title":"Holdom, Two U(1)’s and Epsilon Charge Shifts, Phys","cited_arxiv_id":null,"evidence_quote":"introduces the kinetic-mixing term that defines the dark photon's coupling to the electromagnetic current."},{"cited_title":"Dark $Z$ Implication for Flavor Physics","cited_arxiv_id":"1504.07415","evidence_quote":"earlier dark-Z flavor calculation whose missing q2-independent monopole contribution this paper corrects."},{"cited_title":"A Mathematica Package for Calculation of One-Loop Penguins in FCNC Processes","cited_arxiv_id":"1311.5546","evidence_quote":"supplies the one-loop penguin amplitudes for the b→s, s→d, and b→d FCNC transitions."},{"cited_title":"Search for $B^{+} \\to K^{+} \\nu \\bar \\nu$ decays with an inclusive tagging method at the Belle II experiment","cited_arxiv_id":"2105.05754","evidence_quote":"sets the B+→K+νν upper bound used to check the on-shell ZD→invisible channel."}],"review_version":1}