{"id":"808962d2-6942-47f2-b573-732343d43ca5","arxiv_id":"1908.11867","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"Modular invariant lepton models at levels 4 and 5, using flavons for charged leptons and a single modulus for neutrinos, fit the data with five parameters and predict absolute neutrino masses and CP phases.","lead":"The paper builds new models of lepton masses where neutrino masses are controlled by one modular field and charged lepton masses by separate flavon fields. It shows the models can fit current neutrino data with few parameters and predict neutrino masses, CP-violating phases, and a double-beta decay rate.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 5WC3 best-fit point cannot produce the electron mass: determinant/Frobenius bound forces the smallest singular value of Ye above 1.8e-5, six times y_e.","rationale":"The reader's conditional verdict focused on the assumed flavon vacuum alignment, which is a legitimate model-building caveat but not a correctness flaw in the existence claim. My concern is different and more concrete: the printed numerical inputs for the two level-5 Weinberg scenarios appear unable to reproduce the electron Yukawa coupling, because the determinant of the charged-lepton Yukawa matrix at the quoted point is too large relative to its Frobenius norm. This is an internal, checkable inconsistency in the paper's headline level-5 results. The bound is elementary and does not depend on scanning or global minima: any matrix with Frobenius norm squared ~9.8e-5 and determinant ~8.8e-10 has smallest singular value at least ~1.8e-5, whereas the fitted electron Yukawa is 2.8e-6. If this check confirms, the 5WC3 and 5WC3p fit points are not valid, and the paper's claim that the level-5 Weinberg models give excellent agreement with data is unsupported. Level 4 models and the 5SC seesaw case may still be viable, so I do not reject the whole framework; but the manuscript should not be accepted on the strength of the level-5 Weinberg fits until the singular-value check is performed and the tables are corrected or the fits re-derived. The verdict is therefore unverified rather than conditionally acceptable on the stated basis.","tokens_in":20146,"tokens_out":45561,"duration_ms":378165,"concrete_test":"Recompute the singular values of the 3x3 matrix Ye in Eq. (24) at the Table 8 best-fit points, using the same cos(beta) convention as the table caption, for 5WC3, 5WC3p, and 5SC. The smallest singular value must equal y_e(mZ)=2.794745e-6 (and the determinant must equal y_e y_mu y_tau ~ 1.65e-11 up to the stated convention). For 5WC3, check whether |det(Ye)| <= (||Ye||_F^2/2) y_e ~ 1.37e-10; the printed values give |det(Ye)| ~ 8.8e-10. If the authors' unrounded parameters still violate this bound, the level-5 Weinberg fits cannot reproduce the charged-lepton masses as stated.","verdict_should_be":"UNVERDICTED","load_bearing_attack":"In the level-5 Weinberg case 5WC3, the paper's best fit (chi2_min ~ 1.1), the charged-lepton sector appears internally inconsistent. Using Eq. (24) with the Table 8 inputs (alpha=3.018e-3, beta=3.927e-3, gamma=-0.4484e-3, phi2=0.4260, phi3=0.8030), one obtains ||Ye||_F^2 ~ 9.8e-5 and det(Ye) ~ 8.8e-10. For any 3x3 matrix, det(Ye) = s1 s2 s3 and s1 s2 <= ||Ye||_F^2/2, so the smallest singular value obeys s3 >= 2|det(Ye)|/||Ye||_F^2 ~ 1.8e-5. The electron Yukawa used in the fit is y_e(mZ)=2.794745e-6, about 6.5 times smaller. Thus the printed 5WC3 point cannot reproduce the electron mass, regardless of the unitary rotations that diagonalize the Yukawa matrix. The same bound gives s3 >= 6.8e-6 for 5WC3p, also incompatible with y_e. The seesaw case 5SC has |det| ~ 2.9e-11 and may be consistent, so the problem is specific to the two Weinberg level-5 scenarios that the paper highlights. If the quoted alpha,beta,gamma are raw superpotential parameters with a separate cos(beta) factor, the physical matrix is scaled by cos(beta) <= 1, which only lowers the largest singular value below y_tau and cannot rescue the determinant mismatch. This is a checkable internal-consistency issue in the central numerical claim, not a question of vacuum alignment or naturalness.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes modular-invariant lepton models at levels N=4 (S4) and N=5 (A5) in which charged-lepton Yukawa couplings are built from flavon VEVs while neutrino masses depend on the modulus alone, through either the Weinberg operator or type-I seesaw. Seven scenarios are selected and fitted to the six neutrino observables plus the three charged-lepton Yukawas; the paper reports chi-squared minima, pulls, and predictions for absolute neutrino masses, Majorana phases, the Dirac phase delta, and neutrinoless double beta decay. The level-4 construction reproduces the e/mu/tau hierarchy through modular weights, and the level-5 construction places left- and right-handed charged leptons in A5 triplets. The authors explicitly treat flavon VEVs as free parameters and do not attempt to stabilize them dynamically.","tokens_in":20708,"tokens_out":28145,"duration_ms":262287,"significance":"The paper is clearly written and the group-theoretic setup is explicit: modular-form bases, generators, and Clebsch-Gordan coefficients are collected in appendices, and the numerical fits are reported with pulls and chi-squared values. If the fits are correct, the paper would provide a useful proof of principle that flavons plus modular invariance can generate charged-lepton hierarchies without strong hierarchies in the Lagrangian parameters, with falsifiable predictions for m1, m2, m3, phases, and m_ee. However, the central level-5 Weinberg best-fit point fails an elementary singular-value consistency check, so the numerical claims as printed cannot be accepted without correction.","major_comments":[{"comment":"The 5WC3 best-fit point cannot reproduce the charged-lepton spectrum. Using Eq. (24) with the Table 8 inputs alpha=3.018e-3, beta=3.927e-3, gamma=-0.4484e-3, phi2=0.4260, phi3=0.8030, one obtains ||Y_e||_F^2 ~ 9.8e-5 and |det Y_e| ~ 8.8e-10. Since det(Y_e)=s1 s2 s3 and s1 s2 <= ||Y_e||_F^2/2, the smallest singular value obeys s3 >= 2|det Y_e|/||Y_e||_F^2 ~ 1.8e-5, whereas the electron Yukawa used in the fit is y_e=2.794745e-6. No bi-unitary rotation can remove this discrepancy. Moreover, the Frobenius norm at this point is ||Y_e||_F ~ 9.9e-3, below y_tau=1.003e-2, so the largest singular value is also insufficient for the tau mass. A common rescaling by cos(beta) cannot cure both problems: it lowers the largest singular value further and cannot increase it toward y_tau. The analogous check for 5WC3p appears to give a smallest-singular-value bound above y_e as well, while the 5SC point may satisfy the bound. The statement in Section 3.1 that charged-lepton pulls are negligible is therefore contradicted by the printed parameters; please refit or correct these points and report the actual singular values of Y_e for every scenario.","section":"Section 3.3, Table 8, Eq. (24)"},{"comment":"The numerical results are not reproducible as reported. The minimization algorithm, starting points, and tolerances are not specified, and no uncertainties are given for the fitted parameters or for the predicted quantities (m_i, Majorana phases, m_ee). Because the abstract claims predictions for these quantities, the paper should provide at least a covariance matrix or parameter ranges, together with the charged-lepton Yukawa pulls (or singular values of Y_e) for each of the seven best-fit points. At present, the only quantitative support for fit quality is the quoted chi-squared minimum, and for the 5WC3 case that support is invalidated by the inconsistency documented in the previous major comment.","section":"Section 3.1"}],"minor_comments":[{"comment":"The middle panels list units for m1, m2, m3, and m_ee as 'eV^-2' and 'eV^-1', which are typos for 'eV'; please correct these units.","section":"Tables 7 and 8"},{"comment":"The second and third rows of the Yukawa matrix are typeset in a way that obscures which entries carry superscripts; please retypeset the matrix unambiguously, since the singular-value consistency check in the major comments depends on its precise form.","section":"Eq. (24)"},{"comment":"The paper explicitly states that no dynamical mechanism selects the flavon VEVs; because all charged-lepton results depend on these alignments, the conclusions should state more prominently that the scenarios are effective examples conditional on assumed vacuum alignments rather than complete models.","section":"Sections 2 and 4"},{"comment":"The level-5 cases are labeled 'CP modified' in Table 5, while the text says they have a CP-conserving Lagrangian with real parameters; please clarify in the table caption the difference between a CP-conserving Lagrangian and spontaneous CP violation by the modulus and flavon VEVs.","section":"Table 5 and Section 3.3"}],"recommendation":"major_revision","confidential_remarks":"The paper's main numerical claim for the level-5 Weinberg models fails an elementary singular-value consistency check with the printed parameters. If the authors can produce a corrected fit, the paper may be publishable after revision; if the level-5 Weinberg points cannot be refitted with the charged-lepton masses enforced, those sections should be withdrawn or the conclusions changed. The rest of the construction (level-4 scenarios and the level-5 seesaw case) appears more robust but still lacks parameter uncertainties and charged-lepton pulls."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nHere's the quick version: the paper's headline level-5 Weinberg scenarios (5WC3 and 5WC3p) have a serious internal inconsistency. Using the best-fit parameters in Table 8 and Eq. (24), the charged-lepton Yukawa matrix Ye for 5WC3 has Frobenius norm squared ~9.8e-5 and determinant ~8.8e-10, so its smallest singular value is at least 1.8e-5. That's about six times y_e(mZ)=2.795e-6. The 5WC3p point gives a bound of 6.8e-6, also too large. No unitary rotation can fix this: the electron mass is a singular value of the same matrix. The printed fit points simply do not reproduce the electron mass, despite the claim that charged leptons are reproduced with negligible pulls. This is a load-bearing numerical error, not a matter of vacuum alignment or naturalness.\n\nWhat is genuinely good: the level-4 construction, where differences in modular weights play the role of Froggatt-Nielsen charges and a single flavon VEV phi2=0.01 produces the mass hierarchy with order-one coefficients, is clean and works. Those fits (4WV, 4SV, 4WC, 4SC) pass the determinant sanity check by construction since Ye is diagonal. The level-5 idea of treating left- and right-handed leptons as A5 triplets is interesting, and the seesaw case 5SC, predicting m1=0 and normal ordering, may still be consistent—the stress-test finds no determinant obstruction there. The paper is transparent about what it does not do (no dynamical vacuum selection), and the appendices give the group theory and q-expansions.\n\nThe soft spots beyond the flaw: no minimization algorithm, no parameter error bars, no code; only a curated set of seven scenarios is shown. The lack of reported charged-lepton pulls is what let the 5WC3 inconsistency hide. And the flavon alignments are assumed, not derived—a common but real limitation.\n\nBottom line: the paper deserves a serious referee, but the level-5 Weinberg claims need to be corrected or withdrawn. The level-4 part stands on its own. I would send it to review, with a request for the numerical fitting code or at least a verification of every reported minimum against the charged-lepton Yukawas.","headline":"The level-5 Weinberg fits fail an internal consistency check, but the level-4 construction is a solid and citable piece of model building.","tokens_in":21202,"tokens_out":5443,"would_cite":false,"duration_ms":46847,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["11.30.Hv","12.15.Ff","14.60.Pq"],"model":"deepseek-v4-flash","headline":"Modular invariant models at levels 4 and 5, with flavons in the charged-lepton sector, fit all measured lepton masses and mixing angles with five free parameters and predict absolute neutrino masses, mass ordering, and CP phases.","keywords":["modular invariance","lepton masses and mixing","flavons","finite modular groups","neutrino mass predictions","CP violation","Weinberg operator","type I seesaw"],"falsifier":"Compute the scalar potential of the modulus plus flavons and check whether the vacuum alignments used in the fits, for example $\\phi\\propto(0,0.01,0)$ at level 4 and real $\\phi_2,\\phi_3$ at level 5, are stationary points; if no such minima exist, the charged-lepton matrices no longer take the fitted form and the scenarios are invalid. On the experimental side, the models are settled by the absolute mass scale and by $m_{ee}$: a measurement that excludes a lightest neutrino mass near 40 meV, or that excludes $m_{ee}\\simeq 60$ meV (level-4 Weinberg), $m_{ee}\\simeq 40$ meV (level-4 seesaw), $m_{ee}\\simeq 27$ meV (level-5 Weinberg) and $m_{ee}\\simeq 1.3$ meV (level-5 seesaw), would rule out the corresponding scenarios.","tokens_in":19959,"feed_emoji":"⚛️","tokens_out":15051,"duration_ms":120904,"temperature":0.7,"pith_summary":"The paper sets out to show that the charged-lepton mass hierarchy can be explained inside the modular-invariance framework without fine-tuning, by adding ordinary flavon fields to the modulus. In the models built at levels 4 and 5, neutrino masses and mixing are controlled by the modulus alone, while electron, muon and tau masses come from flavon vacuum values, with modular weights playing the role of Froggatt-Nielsen charges at level 4. The authors fit the model parameters to measured charged-lepton Yukawa couplings and neutrino oscillation data and report seven scenarios with chi-squared values between about 0.3 and 12.6, together with predictions for the absolute neutrino masses, the mass ordering, the Majorana phases and the neutrinoless double beta decay parameter. If the central claim is right, modular invariance can describe the whole lepton sector with a modest number of free parameters rather than being confined to the neutrino sector.","feed_headline":"Five neutrino parameters fit all measured lepton data","feed_subtitle":"Level-4 and level-5 models predict neutrino masses and CP phases testable by neutrinoless double beta decay.","key_machinery":"The central mechanism is a division of labour between the modulus and two kinds of fields. The modulus $\\tau$ alone controls the neutrino sector, through the weight-2 modular forms of levels 4 and 5, while the charged-lepton Yukawa matrices are built only from the vacuum values of ordinary flavons, chiral superfields that are gauge singlets but carry nontrivial representations and weights of the finite modular group ($S_4$ at level 4, $A_5$ at level 5). The weights fix which powers of the flavons appear, so at level 4 they act as Froggatt-Nielsen charges and generate the charged-lepton hierarchy with comparable coefficients; at level 5 the same setup lets right-handed charged leptons sit in the same type of triplet as their left-handed partners. Keeping the charged-lepton sector flavon-only is what allows the neutrino sector to stay minimal.","core_discovery":"The paper's central claim is that ordinary flavons can carry the charged-lepton sector in modular invariant models without spoiling the predictive power of the neutrino sector. At level 4 ($\\Gamma_4\\cong S_4$), giving the right-handed charged leptons different modular weights and letting the flavon enter through powers fixed by those weights produces the electron-muon-tau hierarchy with comparable-size coefficients; at level 5 ($\\Gamma_5\\cong A_5$), left- and right-handed charged leptons are both assigned to irreducible triplets and the charged-lepton Yukawa matrix depends on two flavon vacuum values. Neutrino masses are generated either by the Weinberg operator or by type I seesaw and depend only on the modulus, the overall scale and one parameter $\\xi$, leaving five free parameters in the neutrino sector. The paper reports seven scenarios with $\\chi^2_{\\rm min}$ values between about 0.3 and 12.6 and derives predictions that were not inputs: nearly degenerate neutrino spectra with a lightest mass near 40 meV at level 4, inverted ordering for Weinberg cases and normal ordering for seesaw cases, $m_{ee}\\simeq 60$ meV (Weinberg) or 40 meV (seesaw) at level 4, $m_{ee}\\simeq 27$ meV in the best level-5 Weinberg case, and a massless lightest neutrino with $m_{ee}\\simeq 1.3$ meV in the level-5 seesaw case.","pith_inferences":[],"forward_implications":["The charged-lepton mass hierarchy can be produced by modular weights plus flavon vacuum values with comparable-size Lagrangian couplings, so the hierarchy does not need to be put into the Yukawa couplings by hand.","The level-4 models predict a nearly degenerate neutrino spectrum with the lightest neutrino around 40 meV, so a measurement of the absolute mass scale in this range would support the construction.","The pattern that the Weinberg operator gives inverted ordering and type I seesaw gives normal ordering (with one poor-fit exception) means a definitive determination of the mass ordering will distinguish between the two neutrino-mass mechanisms within this framework.","The predictions for $m_{ee}$, roughly 40-60 meV at level 4, about 27 meV in the best level-5 Weinberg case, and about 1.3 meV in the level-5 seesaw case, put the scenarios within reach of next-generation neutrinoless double beta decay searches.","CP can be conserved by the Lagrangian and still produce large observable CP violation through the vacuum values of $\\tau$ and the flavons, as in the level-5 Weinberg case with $\\delta/\\pi\\simeq 1.7$.","Editorial inference: the paper's division of labour suggests a natural next step, assigning each fermion sector its own modulus so the charged-lepton hierarchy would come from a second modulus rather than from hand-set flavon alignments; the authors state their examples are a first step in this direction.","Editorial inference: if future data fix normal ordering and exclude $m_{ee}$ above about 10 meV, the Weinberg-based scenarios, all predicting inverted ordering, would be eliminated and only the seesaw variants of this construction would survive.","Editorial inference: the level-5 seesaw prediction of a strictly massless lightest neutrino is a sharp, testable signature; a normal-ordered spectrum with a measured nonzero smallest mass would exclude that specific scenario."],"supporting_citations":[{"why":"Supplies the modular-invariance formalism for lepton masses and mixing that the paper builds on, defining how Yukawa couplings are modular forms of the modulus.","marker":"[10]"},{"why":"Gives the consistent implementation of CP transformations in modular invariant theories, used here to impose real Lagrangian parameters in the CP-conserving scenarios.","marker":"[11]"},{"why":"Provides a previous level-3 model with flavons in the charged-lepton sector, the approach the paper extends to levels 4 and 5.","marker":"[44]"},{"why":"Supplies the five independent weight-2 modular forms of level 4 and their $S_4$ transformation properties, which enter the neutrino mass matrices.","marker":"[48]"},{"why":"Supplies the weight-2 modular forms of level 5 that build the $\\Gamma_5$ neutrino mass matrices.","marker":"[51]"},{"why":"Provides level-5 modular form constructions and earlier level-5 models with diagonal charged-lepton sectors, the baseline from which the paper's non-diagonal $A_5$ treatment departs.","marker":"[52]"},{"why":"Provides the renormalized charged-lepton Yukawa couplings at the $m_Z$ scale that the fit uses as experimental targets.","marker":"[59]"},{"why":"Provides the neutrino oscillation data, mass-squared differences, mixing angles and the CP phase, with errors used in the $\\chi^2$ fit.","marker":"[60]"},{"why":"Supplies the explicit $A_5$ representation matrices and Clebsch-Gordan coefficients used to write the level-5 invariants.","marker":"[62]"}],"fun_headline_variants":["Modular weights encode lepton mass hierarchy","Five parameters predict neutrino masses and CP phases","Level-4 and -5 models fit lepton data","Flavon weights explain charged lepton hierarchy","Neutrino predictions from modular symmetry at levels 4 and 5"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the flavon fields can be put at the specific values the fits need, for instance $\\phi\\propto(0,0.01,0)$ at level 4, even though the model contains no mechanism that would select those values; should a more complete theory force different alignments, all the fitted mass matrices and predictions would no longer follow.","fun_headline_variants_meta":{"raw":{"variants":["Modular weights encode lepton mass hierarchy","Five parameters predict neutrino masses and CP phases","Level-4 and -5 models fit lepton data","Flavon weights explain charged lepton hierarchy","Neutrino predictions from modular symmetry at levels 4 and 5"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000617,"raw_usage":{"total_tokens":2930,"prompt_tokens":1075,"completion_tokens":1855,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":691,"completion_tokens_details":{"reasoning_tokens":1779}},"tokens_in":691,"tokens_out":1855,"duration_ms":12440,"temperature":1.0,"reasoning_tokens":1779,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T10:06:22.194731+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the scalar potential of the modulus plus flavons and check whether the vacuum alignments used in the fits, for example $\\phi\\propto(0,0.01,0)$ at level 4 and real $\\phi_2,\\phi_3$ at level 5, are stationary points; if no such minima exist, the charged-lepton matrices no longer take the fitted form and the scenarios are invalid. On the experimental side, the models are settled by the absolute mass scale and by $m_{ee}$: a measurement that excludes a lightest neutrino mass near 40 meV, or that excludes $m_{ee}\\simeq 60$ meV (level-4 Weinberg), $m_{ee}\\simeq 40$ meV (level-4 seesaw), $m_{ee}\\simeq 27$ meV (level-5 Weinberg) and $m_{ee}\\simeq 1.3$ meV (level-5 seesaw), would rule out the corresponding scenarios.","supporting_citations":[],"review_version":1}