{"id":"eeed42aa-e076-4b69-8ae9-a32b21770ae2","arxiv_id":"2608.10695","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"high","formal_verification":"none","parameter_count":8,"one_line_summary":"The authors compute helium-like S-state energies with a Lagrange mesh and fit a 9-parameter Pade-type formula to them, claiming 10-14 digit accuracy.","lead":"A numerical grid method produces very precise energies for two-electron ions, and a nine-parameter formula is fitted to reproduce them across nuclear charges. The formula is accurate on the data it was fitted to, but independent validation is missing, so the headline accuracy is not yet established.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Global accuracy claim rests entirely on in-sample fit; no held-out Z values are used to test the nine-parameter Padé interpolant, so 10-14 s.d. for 'any Z' is unsubstantiated.","rationale":"The reader's weakest assumption is exactly the load-bearing point: the interpolation formula is checked only on the same Z values used for fitting, so the 10\\u201314 s.d. claim is a goodness-of-fit statement rather than a validated prediction. I agree with that assessment. The additional internal inconsistency in the claimed reproduction of Puiseux coefficients (Tables V\\u2013VI versus Eqs. (7)\\u2013(8)) strengthens the concern that the ansatz is not being tested against independent constraints. None of this proves the formula is wrong, so rejection is too strong; the appropriate outcome remains CONDITIONAL pending a held-out validation and preferably an independent high-Z benchmark. The reader's verdict therefore needs no change.","tokens_in":14704,"tokens_out":6902,"duration_ms":77014,"concrete_test":"Refit Eq. (13)/(16) for each of the four states using only a subset of the LMM energies (e.g., Z = 1, 2, 4, 6, 8, 10, 30, 50) and compare against the held-out LMM energies at Z = 3, 5, 7, 9, 20, 40 and at every integer Z in 11\\u201319, 21\\u201329, 31\\u201339, 41\\u201349. If any held-out energy deviates from the formula by a relative error above 10^-10 (i.e., fewer than 10 s.d.), the 'any Z' accuracy claim is not supported; if all agree to 10+ s.d., the in-sample objection is answered.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that the nine-parameter two-point Padé approximant (13)/(16) reproduces 10\\u201314 s.d. for all physically relevant Z. The only supporting comparison is Tables I\\u2013IV, which compare the formula with the same LMM energies used to determine its nine free parameters: \\u00a7IV.A states that the parameters are fixed by fitting the LMM data in Table I, and Table I contains every Z shown. This is an in-sample goodness-of-fit, not a predictive test. With nine parameters and roughly thirteen to fourteen tabulated points per state, agreement at the fitted nodes is expected and does not establish accuracy at unlisted integer Z (e.g., 15, 25, 35, 45) or outside the fitted range. The theoretical matching also appears internally inconsistent: the text says p1 = \\u22121.12349 is correctly reproduced for 2^1S, but Table V lists \\u22120.843; for the triplet states, Eq. (8) gives q1 = \\u00b10.011065 while Table VI lists \\u22120.0210 and \\u22120.009. Thus the claimed global accuracy is not independently established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports Lagrange-mesh method (LMM) calculations for four low-lying S-states of helium-like ions, 1^1S, 2^1S, 2^3S, and 3^3S, for Z up to 50, and claims 14-15 significant digits for this numerical method. It then constructs a generalized two-point Padé approximant in the variable lambda = sqrt(Z - Z_B), matching the large-Z 1/Z expansion and the small-|Z - Z_B| Puiseux expansion around a second critical charge Z_B. For each state the approximant is stated to have nine free parameters fitted to LMM data, and the paper claims 10-14 significant digits for any physically relevant Z, with absolute accuracy as Z -> infinity. Explicit parameters and comparison tables are provided for all four states.","tokens_in":14996,"tokens_out":4979,"duration_ms":51884,"significance":"If the accuracy claims hold, the paper would be a useful contribution: it provides an explicit, cheap interpolation formula that could replace separate high-precision variational calculations for each Z, and it offers LMM data for Z = 20-50 where the authors state no accurate comparison results exist. The tables are explicit, the method is reproducible, and the LMM results for Z <= 10 agree with independent high-precision references to roughly the claimed number of digits. However, the central interpolation claim is currently validated only in-sample, and there are internal inconsistencies between the text and the parameter tables that must be resolved before the claimed accuracy can be accepted.","major_comments":[{"comment":"The central claim that the nine-parameter Padé approximant (13)/(16) reproduces 10-14 s.d. for 'any physically relevant Z' is not independently established. Section IV.A states that the parameters are fixed by fitting the LMM-based data in Table I, and Tables I-IV contain exactly the Z values shown in the accuracy comparison. Thus the agreement is an in-sample goodness-of-fit, not a predictive test. No held-out Z values (for example Z = 15, 25, 35, 45) and no cross-validation are reported, and for Z > 10 there is no independent high-precision data set used for comparison. Please provide an out-of-sample test or at least a residual/convergence analysis showing that the fitted form does not merely interpolate its training points.","section":"§IV.A, Tables I-IV"},{"comment":"The text's claim that the interpolant reproduces the Puiseux coefficients is contradicted by the tables. Section IV.B says the coefficient p1 = -1.12349 is correctly reproduced for 2^1S, but Table V lists p1 = -0.843 for 2^1S. Similarly, Sections IV.C and IV.D say q1 = ±0.011065 and p1 = -0.884233 from Eq. (8) are reproduced, but Table VI lists q1 = -0.0210 and -0.009 and p1 = -0.8091 and -0.8736 for the two triplet states. This discrepancy is load-bearing because the Puiseux matching is the stated theoretical basis of the interpolation. Please clarify which values are outputs of the unconstrained fit and which are imposed constraints, and correct the text or the tables.","section":"§IV.B, Table V; §IV.C-D, Table VI"},{"comment":"The triplet-state input is not independently verifiable. The values Z_B = 0.8799, E_B = -0.384708, and the Puiseux coefficients in Eq. (8) are said to come from the unpublished work in progress [18], and they are obtained by fitting the same LMM data that is later used to assess the interpolation. The paper provides no derivation or external evidence for these quantities. Since the small-Z behavior of the Padé approximant depends on them, this is a load-bearing unverified input. Please include the supporting analysis or a reference where it is published.","section":"§II, Eq. (8), ref. [18]"},{"comment":"The LMM accuracy claim of 14-15 s.d. for Z = 20-50 is not supported by any external comparison. The tables mark these rows as 'first calculation' with no accurate reference values, and the paper reports no convergence study with respect to lattice size for these Z values. The interpolation accuracy statement inherits this uncertainty, because the Padé parameters are fitted to these LMM numbers. Please add lattice-convergence evidence for large Z or explicitly qualify the accuracy claim as an internal LMM estimate.","section":"§I, Tables II-IV (Z = 20-50)"}],"minor_comments":[{"comment":"The phrase 'any physically-relevant nuclear charge Z' overstates the tested range. The tables cover Z = 1 to 50 plus the asymptotic limit; the accuracy at intermediate or larger Z is asserted, not demonstrated. Please qualify the statement to the computed range or provide additional test points.","section":"Abstract and Conclusions"},{"comment":"The parameter-count discussion is confusing: Eq. (11) says the total number of free parameters is 2N+5, then says fixing a0 = E_B reduces it to 2N+4, while Eq. (13) is later said to depend on 9 free parameters with four constraints. Please reconcile these counts in a way that matches the explicit form of Eq. (13).","section":"Eqs. (11)-(13)"},{"comment":"The text says 12 s.d. are reproduced for Z <= 10 and then the accuracy decreases to 9 s.d. at Z = 40, but for Z = 40 the difference between the second and third columns is about 4.7 x 10^-8, which is roughly 10 significant digits in relative terms. The mixing of 'significant digits' and 'decimal digits' should be made consistent.","section":"§IV.B, Table II"},{"comment":"The paper relies on the ground-state Puiseux expansion, yet footnote [24] states that this expansion was never published in complete form. Please provide the complete expansion in an appendix or point to a published source that contains it.","section":"Footnote [24]"},{"comment":"The caption describes the 2^3S state as a red dashed line and the 2^1S state as a gray dashed line, but the plot legend described in the caption may be difficult to distinguish. Please verify the line styles and colors so the figure is unambiguous.","section":"Figure 1 caption"}],"recommendation":"major_revision","confidential_remarks":"The raw LMM data and explicit interpolation parameters are potentially valuable, and the agreement with independent references for Z <= 10 is a genuine strength. The main obstacle is that the headline accuracy claim for the interpolation is validated only on the fitting data, and the mismatches between the text and Tables V-VI for p1 and q1 suggest that the asymptotic-matching claim is not currently reliable. I would want to see a held-out test and a corrected explanation of the Puiseux coefficient reproduction before accepting the paper."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First, the useful part: the paper tables LMM energies for 2^1S, 2^3S, 3^3S up to Z=50, with Z=20-50 genuinely new, and the ground-state column reproduces known high-precision values for Z=2-10. That gives the numerical method some credibility. The two-point Padé interpolation formula, written as a rational function of sqrt(Z-ZB), is a reasonable and compact way to represent the Z-dependence, and extending it to excited states is a natural next step.\n\nThe soft spot is the validation logic. The nine parameters are fitted to the LMM data in Tables I-IV, and the claimed 10-14 s.d. accuracy is then demonstrated by comparing the formula to the same tables. That is a goodness-of-fit statement, not a prediction. No held-out Z values, no independent high-Z comparison, no convergence study, and no code or data files are provided. As a result, \"any physically-relevant nuclear charge\" is not supported by the evidence in the paper.\n\nThere is also an internal inconsistency that needs fixing. The text says the coefficient p1=-1.12349 for 2^1S is reproduced accurately, but Table V lists -0.843. For the triplets, Eq. (8) gives q1=±0.011065 while Table VI lists -0.0210 and -0.009. Either the tables or the text are wrong, and a referee needs to know which. The Puiseux constraints for triplets also rely on an unpublished manuscript [18], and the ground-state expansion is acknowledged in footnote [24] as never published in complete form.\n\nThis doesn't mean the LMM numbers are bad or the interpolation idea is worthless. The high-Z excited-state tables could be a practical resource, and the formulas, once validated on held-out data, could replace expensive per-Z variational calculations for some purposes. But the central accuracy claim, as written, is not established.\n\nFor peer review: I would send this to referees. The raw numerical work is plausibly correct and potentially useful, and the flaws are fixable rather than fatal. A serious referee should ask for the LMM data, a fit/validate split, and a cleanup of the coefficient mismatches. I would not cite the 10-14 s.d. claim in my own work until that happens.","headline":"The excited-state LMM tables are a plausible new dataset, but the 10-14 s.d. interpolation claim is unsupported: it is validated only in-sample and the paper contains internal coefficient mismatches.","tokens_in":15648,"tokens_out":3575,"would_cite":false,"duration_ms":37426,"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":"This paper claims that one nine-parameter rational function reproduces the lowest S-state energies of helium-like ions for all relevant $Z$, with the Lagrange mesh method supplying 14–15 digit benchmarks.","keywords":["helium-like ions","Lagrange mesh method","two-point Padé approximant","second critical charge","Puiseux expansion","1/Z expansion","energy interpolation"],"falsifier":"Compute the energy of, say, the $1^1S$ state at $Z = 15$ or $Z = 35$ with an independent high-precision variational method to 12 decimal digits and compare with equation (13); disagreement beyond the claimed 10–14 significant digits would falsify the interpolation ansatz. Alternatively, numerically continue the fitted Padé function to complex $\\lambda$ and test whether the nearest singularity to the physical axis is exactly the asserted branch point $Z_B$, and whether increasing the degree $N$ changes the predicted energies by more than the claimed accuracy.","tokens_in":14341,"feed_emoji":"⚛️","tokens_out":25857,"duration_ms":190748,"temperature":0.7,"pith_summary":"This paper tries to establish that the non-relativistic energies of the lowest S-states of helium-like ions — two electrons around a nucleus of charge $Z$ — are given for every physically relevant $Z$ by a single nine-parameter rational function of $\\lambda = (Z - Z_B)^{1/2}$, where $Z_B$ is the second critical charge at which the state ceases to exist. To supply this claim with data, it uses the Lagrange mesh method, which it argues reaches 14–15 significant digits for $Z \\leq 50$ with modest computing effort, including $Z = 20$–$50$ where no accurate references existed before. The picture rests on matching the large-$Z$ $1/Z$ expansion to the Puiseux expansion in half-integer powers around $Z_B$, a branch-point structure previously confirmed for the ground state and here extended to the first excited singlet and two lowest triplet S-states. If correct, the result replaces dozens of separate high-precision calculations with one closed-form energy curve that is exact as $Z \\to \\infty$.","feed_headline":"One rational function reproduces He-like energies to 10-14 digits","feed_subtitle":"Replaces per-Z variational runs: a single formula covers the whole helium-like energy sequence.","key_machinery":"The machinery is a generalized two-point Padé approximant, equation (13): the ratio of a degree-9 polynomial to a degree-5 polynomial in $\\lambda = \\sqrt{Z - Z_B}$. The substitution $\\lambda^2 = Z - Z_B$ converts the Puiseux expansion at the second critical charge into a Taylor series at $\\lambda = 0$ and the $1/Z$ expansion into a Laurent series at infinity, so the rational function interpolates between the two asymptotic regimes. Four constraints (14) impose the exact leading terms: the constant $E_B$ and the vanishing of the $\\lambda$ term for the singlets at $Z_B$, and the exact $\\epsilon_0 Z^2$ and $\\epsilon_1 Z$ coefficients at large $Z$. The remaining nine free parameters are fitted to the Lagrange-mesh energies, and the poles of the approximant are reported to lie away from the physical $\\lambda \\geq 0$ semi-axis, avoiding spurious wiggles in the energy.","core_discovery":"The central claim is that equation (13), the two-point Padé approximant $E(\\lambda) = P_9(\\lambda)/Q_5(\\lambda)$ in $\\lambda = (Z-Z_B)^{1/2}$, reproduces the energies of the $1^1S$, $2^1S$, $2^3S$, and $3^3S$ states to 10–14 significant digits over the entire physically relevant range of $Z$. The interpolant is constrained so that, at $Z = Z_B$, it yields the constant term $E_B$ and, for the singlet states, sets the half-integer term $(Z-Z_B)^{1/2}$ to zero; at large $Z$, it reproduces exactly the leading coefficients $\\epsilon_0$ and $\\epsilon_1$ of the $1/Z$ expansion, so the error vanishes as $Z \\to \\infty$. The remaining nine parameters are fitted to the Lagrange-mesh data of Tables I–IV, and the paper reports at least 12 s.d. agreement for the ground state, 9–12 s.d. for $2^1S$, at least 13 s.d. for $2^3S$, and all 14 s.d. for $3^3S$ across $Z \\in [1,50]$. The paper also extends the second-critical-charge picture: the two singlet states share $Z_B \\approx 0.904854$ and are analytically connected through a square-root branch point, as are the two triplet states at $Z_B \\approx 0.8799$.","pith_inferences":["Because the interpolant is a closed-form rational function, derivatives of the energy with respect to $Z$ are available analytically; the paper does not pursue this, but it could directly yield isotope shifts, polarizabilities, or other $Z$-derivative observables.","The paper fixes $Z_B$ from earlier fits rather than letting it float in the nine-parameter fit; fitting $Z_B$ as a free parameter would test whether the branch-point value is independently determined by the Lagrange-mesh data at the claimed accuracy.","The claimed generality to arbitrary excited states is demonstrated on only four S-states; a direct next test is to apply the same form to a higher Rydberg state or to a P-state, where the branch-point structure is not yet established."],"forward_implications":["The energies of the $1^1S$, $2^1S$, $2^3S$, and $3^3S$ states of any helium-like ion in the non-relativistic static approximation can be obtained from one closed-form rational function to 10–14 significant digits, with the error going to zero as $Z \\to \\infty$.","The Lagrange mesh method with a $50 \\times 50 \\times 40$ lattice gives 14–15 significant digits for these four states at $Z \\leq 50$, filling the previously unknown $Z = 20$–$50$ region with first-time accurate results.","The second-critical-charge picture extends beyond the ground state: the two singlet states share one branch point and the two triplet states share another, linking the states through level crossings in the analytic continuation.","The same two-point Padé interpolation scheme is proposed as a general tool for any excited state of the helium-like sequence and, in principle, for other few-electron atomic systems, provided the critical charge and expansion coefficients are known."],"supporting_citations":[{"why":"establishes the Lagrange-mesh energies and the two-point Padé interpolation for the ground state, including the value of the second critical charge used here.","marker":"[9]"},{"why":"introduces the second critical charge and the idea of expanding the energy in half-integer powers around it, the conceptual basis of the interpolation.","marker":"[10]"},{"why":"provides 41-significant-digit variational ground-state energies for $Z = 1$–$10$ used to validate the Lagrange-mesh numbers.","marker":"[2]"},{"why":"supplies high-accuracy reference energies for the $2^1S$ and $2^3S$ excited states used as benchmarks in Tables II and III.","marker":"[5]"},{"why":"reports that the same second critical charge applies to the $2^1S$ state, grounding the singlet Puiseux expansion.","marker":"[16]"},{"why":"supplies the second critical charges for the lowest two triplet states, used to anchor the triplet interpolations.","marker":"[18]"},{"why":"demonstrates the Lagrange mesh method on the negative hydrogen ion, establishing the numerical accuracy for two-electron systems.","marker":"[6]"},{"why":"is the general review of the Lagrange mesh method whose discretization the numerical calculations use.","marker":"[8]"}],"fun_headline_variants":["One rational fit reproduces He-like energies to 14 digits","Universal formula for helium-like ions: 10-14 digits accuracy","Padé interpolant nails all He-like energy states across Z","He-like energy sequence from a single two-point Padé","Single formula replaces per-Z numerics for helium-like ions"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the true energy of each state traces out the exact smooth nine-parameter curve used for the fit, with no extra singularities nearer than the assumed critical charge; the curve is only checked against the same data points that produced it.","fun_headline_variants_meta":{"raw":{"variants":["One rational fit reproduces He-like energies to 14 digits","Universal formula for helium-like ions: 10-14 digits accuracy","Padé interpolant nails all He-like energy states across Z","He-like energy sequence from a single two-point Padé","Single formula replaces per-Z numerics for helium-like ions"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000822,"raw_usage":{"total_tokens":3704,"prompt_tokens":1157,"completion_tokens":2547,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":773,"completion_tokens_details":{"reasoning_tokens":2462}},"tokens_in":773,"tokens_out":2547,"duration_ms":19769,"temperature":1.0,"reasoning_tokens":2462,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T19:15:14.611566+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the energy of, say, the $1^1S$ state at $Z = 15$ or $Z = 35$ with an independent high-precision variational method to 12 decimal digits and compare with equation (13); disagreement beyond the claimed 10–14 significant digits would falsify the interpolation ansatz. Alternatively, numerically continue the fitted Padé function to complex $\\lambda$ and test whether the nearest singularity to the physical axis is exactly the asserted branch point $Z_B$, and whether increasing the degree $N$ changes the predicted energies by more than the claimed accuracy.","supporting_citations":[{"cited_title":"It has the form of a two-point Pade approximant gPade( N + 4/N)(λ) = PN +4(λ)/QN (λ) of order N as the ratio of two polynomials in the argument λ ≡ (Z − ZB)1/2","cited_arxiv_id":null,"evidence_quote":"establishes the Lagrange-mesh energies and the two-point Padé interpolation for the ground state, including the value of the second critical charge used here."},{"cited_title":"Olivares Pil´ on and A.V","cited_arxiv_id":null,"evidence_quote":"introduces the second critical charge and the idea of expanding the energy in half-integer powers around it, the conceptual basis of the interpolation."},{"cited_title":"Nakashima, H","cited_arxiv_id":null,"evidence_quote":"provides 41-significant-digit variational ground-state energies for $Z = 1$–$10$ used to validate the Lagrange-mesh numbers."},{"cited_title":"Yerokhin, K","cited_arxiv_id":null,"evidence_quote":"supplies high-accuracy reference energies for the $2^1S$ and $2^3S$ excited states used as benchmarks in Tables II and III."},{"cited_title":"Turbiner, W","cited_arxiv_id":null,"evidence_quote":"reports that the same second critical charge applies to the $2^1S$ state, grounding the singlet Puiseux expansion."},{"cited_title":"Drake (Editor), Springer Handbook of Atomic, Molecular, and Optical Physic s, Springer, 2006 (Chapter 11)","cited_arxiv_id":null,"evidence_quote":"supplies the second critical charges for the lowest two triplet states, used to anchor the triplet interpolations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"demonstrates the Lagrange mesh method on the negative hydrogen ion, establishing the numerical accuracy for two-electron systems."},{"cited_title":"Aznabaev, A.K","cited_arxiv_id":null,"evidence_quote":"is the general review of the Lagrange mesh method whose discretization the numerical calculations use."}],"review_version":1}