{"id":"9c6e1942-0cb3-4997-b1c9-1088af3b9d69","arxiv_id":"2512.17235","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"In the Zee model, one-loop EFT running—not direct matching—generates the neutrino mass matrix, and 1% variations in fitted high-scale couplings shift neutrino observables beyond next-generation experimental precision.","lead":"A new EFT calculation tracks how neutrino masses and mixing angles change when quantum corrections between the heavy Zee-model scalars and the weak scale are included. It shows that percent-level changes in the unknown high-energy couplings can shift predicted neutrino parameters beyond the precision of upcoming JUNO, DUNE, and Hyper-Kamiokande measurements, so fits that ignore running will be misleading.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central claim rests on the κij RGE (Eq. 31), taken from [40] plus an unpublished thesis correction; if that β-function is wrong, the running-induced shifts and JUNO comparison in Fig. 9 could change.","rationale":"The paper does a careful EFT calculation: matching conditions are documented, benchmark inputs are tabulated, and the qualitative statement that the Zee-model neutrino mass is log-enhanced and hence running-sensitive is well supported by the leading-log argument in Eq. (39). The central quantitative step, however, is the RGE for κij. All of Fig. 9's sensitivity curves and the 'cancellation spoiled by running' mechanism in Sec. 3.2 are generated by Eq. (31). Because the authors rely on a published result as corrected by an unpublished Honours thesis, and because no code or independent derivation of this particular β-function is included, there is a finite chance that a sign or index error alters the running of κ11 relative to κ12. If so, the shifts could be smaller (or larger) than the JUNO band, changing the paper's headline conclusion. I considered the alternative objection that the S1-S3 scalings are not true full-model degeneracies because they shift charged-lepton masses; the paper explicitly acknowledges this and frames the result as constraining a neutrino-only fit, so this is a limitation, not the load-bearing flaw. The replacement test with the published [40] β-function directly probes materiality of the thesis correction, and would settle the concern without requiring a full re-derivation.","tokens_in":39586,"tokens_out":14536,"duration_ms":148915,"concrete_test":"Rerun the full EFT pipeline for Benchmark 1 with Eq. (31) replaced by the published β-function of Ref. [40] without the [85] corrections, keeping all benchmark inputs and matching conditions fixed; compare the Fig. 9 S1/S2/S3 curves to the JUNO uncertainty band. If the curves remain outside the band, the β-function correction is immaterial to the central quantitative claim; if they move inside, the claim is not robust to this input.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Eq. (31) is the engine of the central claim. The paper's own argument in Sec. 3.2 is that the cancellation between κ11 and the κ12(0)-Yukawa combination at m_H2 is spoiled by the RG running of κ11, so the neutrino mass matrix and the large UV-sensitivity curves of Fig. 9 are direct outputs of this β-function. The authors take Eq. (31) from Ref. [40] as corrected by an unpublished Honours thesis [85]; no independent derivation or public code is provided, and the matching verification described in Sec. 3.1 does not cover this RGE. An O(1) error in any of the Yukawa/trace terms—e.g., the T_{ki}κ_{kj}+T_{kj}κ_{ik} terms or the sign of the g_2 term—would change the running of κ11 relative to κ12, altering both the size and possibly the sign of the shifts in Fig. 9. Since the conclusion is explicitly quantitative (shifts exceed JUNO-era uncertainties), the whole numerical demonstration inherits this single unchecked input.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a two-step effective field theory (EFT) treatment of the Zee model for the hierarchy m_h ≫ m_H2 ≫ m_hSM. It derives 1-loop matching conditions for the Weinberg-like operators when integrating out the charged singlet to the 2HDM EFT and then the second doublet to SMEFT, and combines these with 1-loop RG running in the 2HDM EFT. The central claim is that, after the matching, the entire neutrino mass matrix is generated by the RG running that spoils a cancellation between κ11 and the κ12(0) Yukawa combination, and that percent-level changes in fitted high-scale parameters that leave the full-theory prediction invariant induce shifts in neutrino mass parameters exceeding current and JUNO-era projected uncertainties. Four benchmarks illustrate the effect. The paper does not perform a full parameter scan; the benchmarks are fitted to reproduce the low-energy oscillation data using the full-theory formula and then used to illustrate sensitivity of the EFT calculation.","tokens_in":39978,"tokens_out":4002,"duration_ms":47464,"significance":"If the central results hold, the paper makes a useful and timely point: for a radiative neutrino mass model, a naive 1-loop full-theory calculation may miss large logarithms that are properly resummed only by 1-loop matching plus 1-loop running, and this can affect phenomenological comparisons at the precision of JUNO and future experiments. The paper has several genuine strengths: the matching conditions are nontrivial, presented with diagrams, and stated to have been verified by hand; the two possible hierarchies are both discussed; the role of the charged-lepton Yukawa couplings and trace terms in the running of κ11 is analysed in detail; and the numerical setup includes consistency checks of perturbativity, stability, and the mass hierarchy. The comparison against current and projected JUNO sensitivities in Figs. 9 and 10 is a concrete falsifiable demonstration, even though it is sensitivity-based rather than a full statistical scan.","major_comments":[{"comment":"The engine of the quantitative claim is the 1-loop β-function for the dimension-five coefficients κij. It is taken from Ref. [40] as corrected by the unpublished Honours thesis [85]; no independent derivation, ancillary file, or machine-checkable calculation is provided. The hand-verification mentioned in Sec. 3.1 covers the matching conditions, not this RGE. An O(1) error in any trace or gauge term, e.g. the signs of Tkiκkj or the g2 terms, would change the relative running of κ11 and κ12 and hence the size and sign of the shifts in Figs. 9–10. Because the abstract and Sec. 5 make quantitative statements about exceeding JUNO-era uncertainties, the paper should make the derivation of Eq. (31) fully available, either by reproducing it in the paper or by including a public notebook/appendix that verifies it.","section":"§3.2, Eq. (31) / App. B.1, Eq. (72)"},{"comment":"The authors note that 2-loop RGEs for the dimension-five operators in the general 2HDM EFT are not available and assert that they 'do not expect any issues' with scheme dependence. For a claim that running effects exceed percent-level experimental uncertainties, this expectation needs to be quantified. Concretely, the matching-scale dependence of the final prediction should be tested by varying ¯µ around m_h and m_H2 (e.g. by factors of e^{±1/2}) and showing that the shifts in Fig. 9 are stable, or by estimating the size of the missing 2-loop contributions. Without such a test, the percent-level statement is not established at the claimed precision.","section":"§3, paragraph after Eq. (40)"},{"comment":"The benchmark f and Y2 are fitted so that the full-theory tree-level formula reproduces the same low-energy neutrino parameters that the EFT calculation then shifts. The curves in Figs. 9–10 are therefore sensitivity illustrations along a particular high-scale parameter subspace, not evidence that realistic Zee-model parameter points actually produce these shifts. The text partly acknowledges this ('qualitatively demonstrate', 'will not reproduce quantitatively correct values'), but the abstract and Sec. 5 phrase the conclusion as 'quantum corrections have to be included'. The paper should separate these two statements more sharply and state explicitly that a full prior- or parameter-scan analysis is needed to establish whether the effect is generic in the allowed Zee-model parameter space.","section":"§4.2–4.4, Figs. 9–10"}],"minor_comments":[{"comment":"The wording 'quantum corrections have to be included in studies of neutrino mass parameters' overstates what is demonstrated. The analysis shows that within the chosen benchmarks and fitted subspaces, running effects can exceed experimental uncertainties; it does not show that this happens for all viable Zee-model parameters. Consider phrasing the conclusion as 'can be important' rather than 'have to be included' unless a scan is performed.","section":"Abstract / Sec. 5"},{"comment":"The caption says 'shaded regions represent current and future estimated experimental uncertainties' and that γ increases 'in the direction of the arrows marked', but the arrows are not visible in the arXiv text version and the distinction between current and JUNO-projected bands is not labelled. Please add explicit labels to the figure and mention the different shadings in the caption.","section":"Figure 9 caption"},{"comment":"The table lists m_t = 168.62 GeV while the caption says the inputs are at the scale of the top quark pole mass m_t = 172.4 GeV. The relationship between the two conventions should be stated explicitly to avoid confusion (probably m_t(MS-bar) at the pole-mass scale).","section":"Table 1"},{"comment":"The β-function correction is attributed to an unpublished Honours thesis. If this source remains the only reference for a load-bearing expression, please provide a web link, an institutional repository URL, or include the derivation as a supplementary appendix so that readers can check it.","section":"Reference [85] / footnote 3"},{"comment":"There are several typographical and grammatical slips, e.g. 'We will first in detail the matching procedure' (Sec. 3.1) and 'the hierarchy ... that we eluded to schematically' (Sec. 2.2, should be 'alluded to'). These do not affect the physics but should be corrected.","section":"General copy-editing"}],"recommendation":"major_revision","confidential_remarks":"The paper is a serious EFT analysis and the qualitative message is plausible. My recommendation rests on making the central κij β-function independently verifiable and on quantifying the scheme dependence at the claimed precision. I would encourage the editor to require an ancillary derivation or public notebook for Eq. (31) before acceptance, since the unpublished-thesis source is not sufficient for a load-bearing quantitative claim."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is a solid EFT paper, and the new result is real. Schmidt and Vandeleur give the first full 1-loop matching of the Zee model onto the 2HDM EFT and then onto SMEFT, and they show that the neutrino mass matrix at low energy is generated almost entirely by RG running in the 2HDM EFT, not by the matching itself. Eqs. (23) and (26) are explicit and I see no reason to doubt them; the diagrammatic presentation is clear and the claimed hand verification is plausible. The treatment of both mass hierarchies is a nice touch.\n\nThe benchmarks are honest sensitivity demonstrations. The authors fit f and Y2 to the observed neutrino parameters in the full theory, then show that scalings which leave the full-theory mass matrix invariant produce shifts above JUNO-era uncertainties when the 1-loop RGEs are included. They state this procedure plainly, so I do not count it as a flaw. The cLFV constraints and stability checks are properly included.\n\nThe soft spot is exactly where the stress test points. The whole quantitative claim runs through the κij β-function in Eq. (31), which is taken from Grimus-Lavoura with corrections from an unpublished Honours thesis. The paper itself flags this in footnote 3, but it does not supply an independent derivation or public code. An O(1) error in any of the trace or Yukawa terms would change the size and possibly the sign of the running-induced shifts in Fig. 9. For a claim whose punchline is \"exceeds JUNO precision,\" that is a real gap. This is not fatal — the RGEs are standard and the correction is plausible — but the quantitative conclusion inherits an unchecked input.\n\nTwo smaller points. First, scheme independence is asserted, not demonstrated: the paper notes 2-loop RGEs for dimension-five operators in the 2HDM EFT do not exist and says no issue is expected. That is defensible, but it is an expectation. Second, no code or data files are shipped; for a numerical study of this type, that would help a referee check the benchmark curves.\n\nOverall, the conditional verdict is right. The paper deserves a serious referee. I would send it to review and ask for an independent check of Eq. (31) (or auditable code), and for the abstract/conclusions to soften the predictive claim to what the inherited RGEs actually support.","headline":"A genuinely new 1-loop EFT treatment of the Zee model with reusable matching conditions; the JUNO-comparison plots should be read as sensitivity studies, since the central beta function comes with an uncorroborated unpublished correction.","tokens_in":40403,"tokens_out":2904,"would_cite":true,"duration_ms":31475,"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":"At 1-loop order, the entire Zee-model neutrino mass matrix is generated by RG running inside the two-Higgs-doublet EFT, not by matching at the heavy scale, and percent-level UV rescaling moves predicted oscillation parameters beyond next-ge","keywords":["Zee model","radiative neutrino masses","two-Higgs-doublet effective field theory","renormalization group running","Weinberg operator","neutrino mixing parameters","RG corrections","neutrino oscillation precision"],"falsifier":"Reproduce the paper's numerical procedure with an independent implementation of the 1-loop RGEs for the 2HDM EFT and the two matching steps, apply the same γ scalings of f, Y1_e and Y2_e that preserve the full-theory mass matrix, and compare the resulting displacements with the paper's current and projected experimental bands; a disagreement larger than those bands would indicate an error in the input β-function or matching conditions. A direct two-loop calculation of the Zee-model neutrino mass in the same scheme would settle whether the 1-loop running picture is quantitatively correct.","tokens_in":39470,"feed_emoji":"⚛️","tokens_out":8609,"duration_ms":84824,"temperature":0.7,"pith_summary":"The paper aims to show that a proper effective-field-theory treatment changes how the Zee model produces neutrino masses. When the heavy charged singlet and the second Higgs doublet are integrated out sequentially, the one-loop matching contributions to the dimension-5 Weinberg operator cancel, and what remains is a logarithm of the two heavy scales that is generated by one-loop renormalization-group running in the intervening two-Higgs-doublet EFT. The authors derive the one-loop matching conditions, solve the RG equations numerically for four benchmark sets, and find that percent-level rescalings of the fitted high-scale couplings—rescalings that leave the full-theory neutrino mass matrix invariant—shift the predicted mass-squared differences and mixing angles by more than current and projected experimental uncertainties. They conclude that precision phenomenological studies of the Zee model must include RG corrections. The quantitative size of these effects rests on a 1-loop beta function for the dimension-5 coefficients that the paper takes from the literature as corrected by an unpublished thesis, and the calculation stops at 1-loop order with no 2-loop check.","feed_headline":"Running, not matching, generates Zee-model neutrino masses","feed_subtitle":"Percent-level twists of the high-scale couplings shift predicted neutrino observables beyond next-generation experimental precision.","key_machinery":"The central object is the set of dimension-5 Weinberg-like coefficients κij (i,j = 1,2 for the two Higgs doublets) in the 2HDM EFT; these are the coefficients of the operators (L Hi)(Hj† L) that after electroweak symmetry breaking become Majorana neutrino masses. The paper's argument is carried by the 1-loop β-function for κij (with gauge, Yukawa, and quartic terms), the derived matching conditions from Zee -> 2HDM and 2HDM -> SMEFT, and the cancellation between κ11 and the term κ12(0)Y1e†Y2e + Y2e†Y1eκ12(0) at the intermediate scale. The running of κ11 relative to κ12 is what produces the low-energy neutrino mass matrix.","core_discovery":"Working in the hierarchy m_h > m_H2 ≫ EW scale, the paper matches the Zee model to the two-Higgs-doublet EFT at m_h and then to the SMEFT at m_H2, keeping 1-loop terms. The central observation is that the would-be direct contributions to the SMEFT Weinberg coefficient—κ11 at 1-loop and the tree-level κ12 term combined with charged-lepton Yukawas—cancel exactly at the matching scale, leaving only the log(m_H2^2/m_h^2) term, which is the resummed large logarithm of the full theory. Since the log can only be reproduced by running, the paper states that the entirety of the neutrino mass matrix is generated by RGE running in the 2HDM EFT. In four benchmarks, percent-level variations of UV paramet","pith_inferences":["The split between matching and running is scheme- and matching-scale-dependent; in another scheme part of the effect could be shifted into the matching condition. The scheme-independent content is that a full one-loop EFT calculation, not the naive one-loop full-theory formula, is required for precision.","A direct check of the quantitative claim would be an independent re-derivation of the 1-loop β-function for κij in the general 2HDM EFT; if the coefficients change, the γ-scaling shifts would change and the comparison to experimental precision would need revision.","The flat directions that preserve the full-theory mass matrix could be used in future global fits as a diagnostic: any fitted point that ignores running sits on a degeneracy that the EFT running breaks, effectively constraining the UV parameters at the level of next-generation experiments.","The same two-step EFT construction could be applied to other radiative models, such as Zee-Babu or the scotogenic model, to check whether their neutrino mass matrices likewise arise from running in the intermediate theory."],"forward_implications":["Precision fits of the Zee model should replace the one-loop full-theory formula with a sequential EFT calculation that includes 1-loop matching at m_h and m_H2 and 1-loop RG running between them.","Percent-level freedom in the fitted UV couplings—directions that the full theory does not constrain—changes the predicted neutrino observables by more than current and next-generation experimental errors, so RG corrections must be included when exploiting the new data.","The corrections grow with large quartic coupling λ6, sizable second-doublet Yukawa couplings, and a large hierarchy between the charged singlet and second doublet masses; the same qualitative effect appears for a 1 TeV second doublet.","The large-log approximation reproduces the full-theory formula at leading log, confirming that the running is the source of the logarithm, but in benchmark 1 the β-function varies enough that a constant-β large-log approximation is not quantitatively valid.","The authors expect similarly sized RG corrections in other radiative neutrino mass models, since running enters at the same loop order as the radiative mass itself."],"fun_headline_variants":["Zee neutrino masses come from RGE running alone","Quantum corrections alone forge Zee neutrino masses","Zee model: neutrino mass from running, not matching","RGE running, not matching, yields Zee neutrino masses","All Zee neutrino mass from RGE running alone"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing input is the 1-loop beta function for the dimension-5 coefficients κij in the general two-Higgs-doublet EFT, taken from the published literature as corrected by an unpublished thesis; the cancellation mechanism, the large-log resummation, and every benchmark sensitivity curve inherit this function, and if it is wrong the size of the running effects and the comparison to experimental precision would change.","fun_headline_variants_meta":{"raw":{"variants":["Zee neutrino masses come from RGE running alone","Quantum corrections alone forge Zee neutrino masses","Zee model: neutrino mass from running, not matching","RGE running, not matching, yields Zee neutrino masses","All Zee neutrino mass from RGE running alone"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00068,"raw_usage":{"total_tokens":2860,"prompt_tokens":615,"completion_tokens":2245,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":359,"completion_tokens_details":{"reasoning_tokens":2168}},"tokens_in":359,"tokens_out":2245,"duration_ms":16289,"temperature":1.0,"reasoning_tokens":2168,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T15:20:14.257839+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Reproduce the paper's numerical procedure with an independent implementation of the 1-loop RGEs for the 2HDM EFT and the two matching steps, apply the same γ scalings of f, Y1_e and Y2_e that preserve the full-theory mass matrix, and compare the resulting displacements with the paper's current and projected experimental bands; a disagreement larger than those bands would indicate an error in the input β-function or matching conditions. A direct two-loop calculation of the Zee-model neutrino mass in the same scheme would settle whether the 1-loop running picture is quantitatively correct.","supporting_citations":[],"review_version":1}