{"id":"40720d2a-f5fd-4c20-938b-be899eb96d02","arxiv_id":"2501.12441","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":10,"one_line_summary":"Using CONUS+ reactor data, the authors constrain the weak mixing angle, neutrino electromagnetic properties, and light scalar and vector mediators, finding their most notable limit on neutrino millicharge when electron scattering is included.","lead":"This paper reanalyzes new reactor neutrino data from the CONUS+ experiment to test the Standard Model and search for new physics. It sets limits on neutrino properties and light new particles by combining two types of scattering events.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Eq. (7b) defines a dimensionally inconsistent Q_ℓℓ′; with that printed definition the BSM EνES signal is sub-event-level, so the claimed 10⁻¹² e millicharge limit cannot follow.","rationale":"The reader identified the low-energy EνES modeling and background subtraction as the weakest assumptions. My concern is more specific and more load-bearing: the printed definition of Q_ℓℓ′ in Eq. (7b) is dimensionally inconsistent, and with that definition the EνES millicharge signal in the CONUS+ analysis window is orders of magnitude too small to produce the claimed 10⁻¹² e limit. This is not merely a disagreement with consensus; it is an internal inconsistency that can be checked against the standard photon-exchange derivation. The paper has real strengths: it uses the CONUS+ data release, includes both CEνNS and EνES, treats the correct nuclear form factor, and provides comparisons with other experiments. The weak-mixing-angle and magnetic-moment results are less affected by the Q_ℓℓ′ issue, since the magnetic-moment cross section in Eq. (6) is standard and the mixing-angle analysis is dominated by CEνNS. However, the millicharge and charge-radius constraints, and the headline claim of a four-order-of-magnitude improvement from EνES, rest on Eq. (7b). If the authors used the standard Q in the numerical code, then the printed equation must be corrected and the results re-validated; if they used the printed Q, the quoted limits are unreliable. Either way the paper cannot be accepted as is, but a clear correction path exists, hence CONDITIONAL rather than REJECT.","tokens_in":17828,"tokens_out":58856,"duration_ms":553622,"concrete_test":"Independently compute the EνES event rate for the CONUS+ 1 kg, 119 d, 160 eVee-threshold setup in the lowest 10 eVee bin using (a) the printed Q_ℓℓ′ from Eq. (7b) and (b) the standard photon-exchange Q = 4√2π α (q_ν/e)/(G_F q²), for q_ν = 10⁻¹² e. If (a) gives ≪1 event and (b) gives O(1–10) events, the origin of the 10⁻¹² limit is identified. Also check whether the CEνNS-only limit in Table II (≈10⁻⁸ e) is reproducible with the standard Q; if it instead yields ≈10⁻¹⁰ e, the numerical results are inconsistent with the printed equation.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central millicharge improvement relies on Eq. (7b) and its Q_ℓℓ′ = √(2πα/G_F)(⟨r²⟩/3 − 2q_ν/(e q²)). This quantity is dimensionally inconsistent: √(2πα/G_F) has units of GeV while q² is in GeV², so Q_ℓℓ′ has units GeV⁻¹ and cannot be added to the dimensionless vector coupling g_V. Re-deriving the photon-exchange amplitude gives the dimensionless shift Q = 4√2π α (q_ν/e)/(G_F q²) (and an analogous charge-radius term). At Te = 160 eV, q² = 2m_eTe ≈ 1.64×10⁻¹⁰ GeV², so for q_ν/e = 10⁻¹² the printed Q ≈ 0.77, whereas the standard Q ≈ 68. With the printed Q, the EνES millicharge signal in the CONUS+ 160–320 eVee window is only ~10⁻² events (the SM EνES rate in that window is already ~10⁻² events), so it cannot move the CEνNS-only limit of ~10⁻⁸ e to the claimed ~10⁻¹² e. The quoted limit requires Q ~ 20–100, i.e. the standard normalization. This internal inconsistency directly affects the millicharge and charge-radius constraints in Tables II–III and the paper's headline claim of a four-order-of-magnitude improvement from including EνES.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper reanalyzes the CONUS+ reactor antineutrino data, including both CEνNS and EνES signals, to extract constraints on the weak mixing angle, neutrino electromagnetic properties (millicharge, charge radius, magnetic moment), and light vector (B−L) and scalar mediators. The authors report sin^2θ_W = 0.247^{+0.050}_{-0.054}, a combined CEνNS+EνES limit on the effective neutrino magnetic moment of 1.12×10^{-10} μ_B, millicharge limits of order 10^{-12} e (versus 10^{-8} e from CEνNS alone), charge-radius limits comparable to Dresden-II and COHERENT, and a claim that CONUS+ gives the strongest constraint on light scalar mediators for M_ϕ > 6 MeV.","tokens_in":18187,"tokens_out":8444,"duration_ms":81813,"significance":"If the results are correct, the paper makes a striking and useful point: including the subdominant EνES channel in a reactor CEνNS analysis can dramatically sharpen neutrino electromagnetic property limits, especially the millicharge, and can set competitive bounds on light mediators. The analysis uses the public CONUS+ data release, a transparent χ^2 framework with nuisance parameters from the collaboration's quoted systematics, and comparisons with a broad set of existing constraints. The central quantitative claim, however, depends on an equation that as printed is dimensionally inconsistent, so the numerical results are not traceable until that is fixed. The paper also lacks a robustness check of the low-energy EνES atomic-binding model that drives the improved limits.","major_comments":[{"comment":"The definition of Q_ℓℓ′ = √(2πα_EM/G_F) (⟨r²_νℓℓ′⟩/3 − 2/q² · q_νℓℓ′/e) is dimensionally inconsistent. Since √(2πα_EM/G_F) has mass dimension +1 and the bracket has mass dimension −2, Q_ℓℓ′ has mass dimension −1 and cannot be added to the dimensionless couplings g_V and g_A in the same equation. The correct dimensionless shift for the millicharge contribution is proportional to α_EM/(G_F q²) × (q_ν/e), e.g., Q ≈ (4√2π α_EM)/(G_F q²) (q_ν/e), with a separate q²-independent term for the charge radius. Numerically, at T_e = 160 eV and q_ν/e = 10^{-12}, the printed Q ≈ −0.77, whereas the standard dimensionless Q ≈ 68. With the printed Q, the EνES millicharge signal in the lowest recoil bin is of order 10^{-2} events, far too small to move the CEνNS-only limit of order 10^{-8} e to the claimed 10^{-12} e. The authors must correct Eq. (7b) and the definition of Q, provide the proper derivation, and state explicitly which normalization was used to obtain the numerical limits in Tables II and III.","section":"II.B, Eq. (7b)"},{"comment":"The claimed improvement in the millicharge and magnetic-moment limits is driven by the lowest-energy bins (160–320 eVee), where the EνES signal is modeled with a step-function approximation for the effective electron number, Z_eff(T_e) = Σ_j Θ(T_e − B_j), with binding energies from X-ray data. This is a simplified treatment, and the analysis also integrates down to T_e^min = 2.96 eVee before detector smearing. The authors should test the robustness of their results by comparing the step-function model with a more detailed ionization model for germanium, or by varying the Z_eff prescription and demonstrating that the improved limits are stable. Without such a check, the quantitative four-orders-of-magnitude improvement over the CEνNS-only limit is not fully substantiated.","section":"III, Eq. (20) and Fig. 1"}],"minor_comments":[{"comment":"The abstract and introduction state that the detector thresholds are T_th = 160 eVee for C3, 170 eVee for C5, and 180 eVee for C2, but the event simulation in Fig. 1 uses a single effective detector with a 160 eVee threshold; the authors should clarify how the different thresholds are combined in the statistical analysis.","section":"Abstract and I"},{"comment":"The CEνNS-only millicharge limits are given only in the text as {[-0.49, 3.22], ≤1.9, ≤1.9}×10^{-8} e; they should also appear in Table II for completeness and to make the improvement explicit.","section":"IV, Table II"},{"comment":"The effective magnetic moment in Eq. (5) is written for neutrino scattering, but the experiment detects antineutrinos; the authors should specify whether the same expression applies to antineutrinos and whether the CP-conjugate form affects the definition.","section":"II.B, around Eq. (5)"},{"comment":"The conclusions repeat the claim that CONUS+ provides the most stringent scalar-mediator limit for M_ϕ > 6 MeV, but the text does not provide the exact numerical values of the exclusions or the comparison data sources in Fig. 7; adding a short table with the key benchmark points would improve reproducibility.","section":"V"}],"recommendation":"major_revision","confidential_remarks":"The dimensional inconsistency in Eq. (7b) is the central issue. I suspect the numerical analysis used the standard dimensionless Q, and the printed definition is a typographical slip rather than the actual code, because the claimed improvement is consistent with the standard normalization. Nevertheless, the manuscript as written cannot be trusted until the equation is corrected and the results re-derived. The authors should also compare their results with the concurrent work of Ref. [58] (Alpízar-Venegas et al.) and clarify what new aspects their analysis adds beyond that paper."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper is a standard, largely competent analysis of the new CONUS+ CEνNS data with EνES added on. The weak mixing angle, magnetic moment, and light mediator constraints are sensible, and the nuisance treatment tracks the collaboration's quoted systematics. Adding EνES to the BSM scan is a genuine new element, and the authors are upfront about the overlap with Ref. [58].\n\nThe problem is the millicharge and charge-radius part. Eq. (7b) defines Q_ℓℓ′ = √(2πα_EM/G_F)(⟨r²⟩/3 − 2q_ν/(e q²)). This quantity has units of GeV⁻¹ (the prefactor is GeV, the bracket has GeV⁻²), so it cannot be added to the dimensionless g_V. Re-deriving the photon-exchange amplitude gives the dimensionless shift Q = 4√2π α (q_ν/e)/(G_F q²) (plus an analogous charge-radius term). At Te = 160 eV, with q_ν = 10⁻¹² e, the printed Q ≈ 0.8, while the correct Q ≈ 68. That is a factor of roughly 90 in the coupling, which means the EνES millicharge signal in the lowest bins is suppressed by orders of magnitude relative to what the claimed limit requires. The paper's 90% C.L. limit of 1.9×10⁻¹² e and the four-order improvement from including EνES therefore do not follow from the printed formulas. The same issue affects the charge-radius constraints in Table III.\n\nI do not think this is a one-off typo. The claimed result is consistent with the correct normalization, not with the printed one; if the authors used the correct formula in their code, the printed equation is wrong in a way that directly misrepresents the analysis. That is exactly what referees are for.\n\nOther soft spots are secondary: no goodness-of-fit, no code release, and the Zeff step-function model at 2.96 eV might not be reliable enough to support a limit driven by the lowest bin. These matter once the normalization issue is resolved.\n\nBottom line: the paper deserves a serious referee, but the EM-properties section needs a major correction before it can be used as a reference. The mediator and magnetic moment results may survive. I would send it to review, expecting the authors to re-run the millicharge and charge-radius limits with the correct Q and either revise the claim or explain the discrepancy.","headline":"A competent CONUS+ analysis with a load-bearing sign error in the millicharge formula; the headline limit doesn't follow from the paper's own equations.","tokens_in":18744,"tokens_out":10152,"would_cite":false,"duration_ms":96436,"reading_group":"maybe","serious_thinker":"no","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Adding neutrino-electron scattering to reactor CEνNS data tightens neutrino millicharge limits by four orders of magnitude.","keywords":["coherent elastic neutrino-nucleus scattering","elastic neutrino-electron scattering","CONUS+ experiment","neutrino millicharge","neutrino magnetic moment","neutrino charge radius","light mediators","weak mixing angle"],"falsifier":"Compute the same limits with a full atomic many-body calculation of the germanium electron-recoil spectrum in place of the step-function $Z_{\\rm eff}$, and with a covariance matrix spanning the energy bins; if the 90% C.L. interval for $q_{\\nu_e e}$ moves beyond roughly $10^{-12}\\,e$, or the lowest two bins no longer match the predicted rise, the claimed millicharge improvement is an artifact of the low-energy model.","tokens_in":17612,"feed_emoji":"⚛️","tokens_out":8504,"duration_ms":74084,"temperature":0.7,"pith_summary":"This paper reanalyzes the CONUS+ reactor-antineutrino data, which detected coherent elastic neutrino-nucleus scattering (CEνNS) on germanium, and shows that adding the concurrent elastic neutrino-electron scattering (EνES) channel changes what can be concluded from the same exposure. With both channels included, the 90% C.L. bound on the neutrino millicharge tightens from roughly $10^{-8}\\,e$ (CEνNS only) to about $10^{-12}\\,e$, because the lowest-energy events are dominated by EνES. The same combined analysis yields a low-energy weak mixing angle of $\\sin^2\\theta_W = 0.247^{+0.050}_{-0.054}$ and competitive limits on the neutrino magnetic moment, charge radius, and light scalar and vector mediators. If these constraints are correct, CONUS+ becomes the most stringent reactor-based probe of neutrino electromagnetic properties and the strongest current source of scalar-mediator limits for $M_\\phi > 6$ MeV.","feed_headline":"Millicharge limits fall 10,000-fold when electron scattering is fitted","feed_subtitle":"Including neutrino-electron scattering in the CONUS+ fit sets the tightest light-scalar limit above 6 MeV","key_machinery":"The load-bearing object is the combined $\\chi^2$ function of Eq. (21): predicted CEνNS and EνES event rates per energy bin are compared with the CONUS+ background-subtracted reactor-on excess events, with nuisance parameters $\\alpha$ and $\\beta$ absorbing correlated systematics such as flux, quenching, threshold, form factor, and detector mass. The EνES rate is computed with a step-function effective electron number $Z_{\\rm eff}(T_e)=\\sum_j\\Theta(T_e-B_j)$ for germanium atomic binding, and the CEνNS rate uses a Lindhard quenching factor with $k=0.162$ and a Helm form factor. Because millicharge and magnetic-moment contributions grow at low recoil energy, and EνES dominates the lowest bins, including this channel is what multiplies the experiment's sensitivity to neutrino electromagnetic properties.","core_discovery":"The paper's central claim is that the background-subtracted CONUS+ data, when modeled with both CEνNS and EνES, constrains new physics more sharply than a CEνNS-only readout. For the diagonal neutrino millicharge it finds $q_{\\nu_e e}\\in[-1.8,\\,1.9]\\times10^{-12}\\,e$ at 90% C.L., compared with $[-0.49,\\,3.22]\\times10^{-8}\\,e$ without EνES; the transition millicharges satisfy $|q_{\\nu_e\\mu}|,\\,|q_{\\nu_e\\tau}|\\le1.85\\times10^{-12}\\,e$. It also reports $\\mu^{\\rm eff}_{\\nu_e}\\le1.12\\times10^{-10}\\,\\mu_B$, charge-radius bounds $\\langle r^2_{\\nu_e e}\\rangle\\in[-59.76,\\,8.33]\\times10^{-32}\\,\\mathrm{cm}^2$, and exclusion contours for light $B-L$ vector and scalar mediators, with the scalar limit being the most stringent available for $M_\\phi>6$ MeV. Within the Standard Model, it determines $\\sin^2\\theta_W = 0.247^{+0.050}_{-0.054}$ at $1\\sigma$.","pith_inferences":["Because the millicharge limit is driven by the lowest-energy bins, the $10^{-12}\\,e$ scale should be read as contingent on the step-function atomic-binding model and on the background subtraction being unbiased; a dedicated low-energy electron-recoil calibration of germanium would test it directly.","The same combined CEνNS+EνES prescription could be applied to other reactor CEνNS data sets, such as Dresden-II, and to future larger CONUS+ exposure; similar order-of-magnitude gains in millicharge sensitivity may appear there.","This analysis suggests that the practical discovery reach of reactor neutrino experiments for neutrino electromagnetic properties is set less by nuclear-recoil statistics than by the cleanliness and modeling of the sub-keV electron-recoil region."],"forward_implications":["Including EνES in future reactor CEνNS analyses is not optional for BSM searches: it improves the millicharge bound by four orders of magnitude and sharpens the magnetic-moment bound by roughly a factor of four.","CONUS+ alone now gives the most restrictive experimental limit on light scalar mediators for $M_\\phi > 6$ MeV, surpassing existing CEνNS-based limits in that mass range.","The low-energy weak mixing angle extracted from reactor CEνNS, $\\sin^2\\theta_W=0.247^{+0.050}_{-0.054}$, is consistent with Standard Model running and comparable to Dresden-II, though less precise than COHERENT.","The resulting neutrino electromagnetic limits are still weaker than solar-neutrino EνES experiments such as XENONnT, LZ, and Borexino, so the new constraints are complementary rather than world-leading in every channel."],"supporting_citations":[{"why":"Supplies the CONUS+ background-subtracted reactor-on excess event data, detector parameters, and systematic uncertainty inputs used for every fit.","marker":"[6]"},{"why":"Provides the Standard Model elastic neutrino-electron scattering cross section used to model the EνES channel.","marker":"[12]"},{"why":"Provides the Huber reactor antineutrino spectra for fission isotopes, used to compute the neutrino flux above 2 MeV.","marker":"[49]"},{"why":"Provides the Müller parametrization of reactor antineutrino spectra, supplementing the flux model.","marker":"[50]"},{"why":"Gives the measured ionization quenching factor for germanium with $k=0.162$ that converts nuclear recoil energies to electron-equivalent energies.","marker":"[47]"},{"why":"Introduces the step-function effective electron number $Z_{\\rm eff}$ that models atomic binding in the low-energy EνES rate.","marker":"[52]"},{"why":"Supplies the Gaussian chi-square with nuisance parameters that defines the statistical analysis.","marker":"[54]"},{"why":"Provides the combined COHERENT CsI+LAr CEνNS analysis used as the principal comparison baseline for limits and the weak mixing angle.","marker":"[11]"},{"why":"Provides the Dresden-II reactor CEνNS analysis against which the CONUS+ weak mixing angle and electromagnetic-property constraints are compared.","marker":"[55]"}],"fun_headline_variants":["Adding electron scattering to CONUS+ cuts neutrino millicharge bounds 10,000-fold","CONUS+ with electron scattering sets the most stringent light-scalar limit above 6 MeV","Neutrino millicharge bounds tighten 10,000-fold when electron scattering is fitted","Electron scattering inclusion in CONUS+ yields 10,000-fold better millicharge limits"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole analysis rests on treating the CONUS+ background-subtracted reactor-on excess as an unbiased Gaussian residual spectrum and on modeling sub-keV EνES with a step-function atomic-binding approximation, and since the millicharge signal is concentrated in the lowest-energy bins, errors in either would directly change the headline limits.","fun_headline_variants_meta":{"raw":{"variants":["Adding electron scattering to CONUS+ cuts neutrino millicharge bounds 10,000-fold","CONUS+ with electron scattering sets the most stringent light-scalar limit above 6 MeV","Neutrino millicharge bounds tighten 10,000-fold when electron scattering is fitted","Electron scattering inclusion in CONUS+ yields 10,000-fold better millicharge limits"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.002757,"raw_usage":{"total_tokens":10510,"prompt_tokens":950,"completion_tokens":9560,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":566,"completion_tokens_details":{"reasoning_tokens":9463}},"tokens_in":566,"tokens_out":9560,"duration_ms":74357,"temperature":1.0,"reasoning_tokens":9463,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T17:11:53.422983+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the same limits with a full atomic many-body calculation of the germanium electron-recoil spectrum in place of the step-function $Z_{\\rm eff}$, and with a covariance matrix spanning the energy bins; if the 90% C.L. interval for $q_{\\nu_e e}$ moves beyond roughly $10^{-12}\\,e$, or the lowest two bins no longer match the predicted rise, the claimed millicharge improvement is an artifact of the low-energy model.","supporting_citations":[{"cited_title":"Physics implications of recent Dresden-II reactor data","cited_arxiv_id":"2208.13262","evidence_quote":"Provides the Dresden-II reactor CEνNS analysis against which the CONUS+ weak mixing angle and electromagnetic-property constraints are compared."}],"review_version":1}