{"id":"0daca553-5db9-476f-9ee6-4013b7518993","arxiv_id":"2607.02828","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Every even Galerkin vector induces a band-limited Guinand–Weil test function whose zero sum equals the truncated Weil form exactly, and the omitted archimedean tail is a totally positive Cauchy–Stieltjes increment with budget ~ log T / T.","lead":"Finite Galerkin matrices of the truncated Weil form evaluate exact sums over zeta zeros via an explicit band-limited test function, and their archimedean tails obey a totally positive order with an explicit certification budget. This turns Connes-style truncations into calibrated instruments that say what finite numerics can and cannot certify about Weil positivity.","discovery_kind":"extension","skeptic_critique":{"model":"grok-4.5","headline":"No significant objection identified","rationale":"The reader correctly isolates the Guinand–Weil applicability statement (Lemma 2.2 + last paragraph of Theorem 2.5 proof) as the weakest external dependence and still assigns ACCEPT with low correctness risk. That assessment is accurate: the paper supplies an explicit admissible family, matches normalizations entrywise (Lemma 2.1), and never claims more than the classical formula applied to that family. The tail-order argument is self-contained (rank-two density, Andréief + Cauchy determinants, elementary envelope for B_T). Reproducibility artifacts (Arb LDL^T certificates, source-quotient audits through N=30, three independent dictionary routes) further lower risk. No stronger load-bearing concern surfaces; the verdict therefore remains ACCEPT/HIGH.","tokens_in":15734,"tokens_out":544,"duration_ms":13555,"concrete_test":"Independently recompute the worked-example zero-side partial sums of Table 1 (c=13, N=4, pole-neutral v) against the first 512 ordinates using an independent zero list (e.g., Odlyzko or LMFDB) and a separate high-precision quadrature of g_v; confirm that the raw residual still falls to ~10-11 and the tail-corrected residual to ~10-12, matching the closed-form matrix contraction to working precision.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claims (Theorem 2.5 dictionary identity and Theorem 3.2/Corollary 3.3 tail-order certification) rest on elementary closed-form transport of Galerkin vectors to admissible test functions plus classical Cauchy total-positivity applied to an explicit rank-two density. The softest external input is the applicability of the Guinand–Weil explicit formula to the class of Lemma 2.2 (entire of exponential type ≤ L, compact Fourier support in [-Δ, Δ], strip decay O((1+|Re z|)-2)). That class membership is established by two Stieltjes integrations by parts on a compactly supported piecewise-smooth weight that vanishes at the endpoints, which is standard and sufficient for absolute convergence via the Riemann–von Mangoldt local count. No internal inconsistency, hidden assumption, or gap in the finite calculus (source quotient, pole-neutral family, entrywise identification with CCM closed forms) appears; the shipped three-route numerical checks and Arb certificates further corroborate the identities at concrete (c,N).","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"The paper proves two exact finite theorems for the Connes–van Suijlekom and Connes–Consani–Moscovici truncations of the Weil quadratic form at prime cutoff c>1 and band N. First (Theorem 2.5), every real even Galerkin vector v is transported by an explicit chain (trigonometric polynomial → Volterra sine-chord kernel → compact Fourier weight → entire test function g_v) to a band-limited Guinand–Weil test function such that the cutoff-free truncated form equals the sum of g_v over the nontrivial zeros of ζ (with multiplicity); the construction factors through an exact 2N+1-dimensional source quotient and admits a non-collapsing pole-neutral subfamily. Second (Theorem 3.2, Corollary 3.3), the omitted archimedean tail past any T>max(ρN,7) is a rank-two-density Cauchy–Stieltjes increment that is positive definite and strictly totally positive, yielding the two-sided certification rule λ_j(Q_tot_T)<λ_j(Q_∞)≤λ_j(Q_tot_T)+B_T with explicit budget B_T∼(2N+1)ρ log T/(π²T). The dictionary is checked against the first 512 zeros and by three independent computational routes; a full verification package ships with the paper. No RH or prime-location claim is made.","tokens_in":15918,"tokens_out":1087,"duration_ms":25902,"significance":"The results turn the finite Galerkin matrices into calibrated instruments: every quadratic value is an exact zero sum for an explicitly parametrized family of admissible test functions, and finite-cutoff spectra carry elementary, quantitative certification bounds that make precise what numerics can and cannot decide about the cutoff-free form. The closed-form transport (with inverse-free source calculus), the exact 2N+1 quotient, the pole-neutral family of dimension N−s−1, and the strict total-positivity budget are concrete, usable contributions to the spectral approach to Weil positivity. The released Arb interval certificates, three-route dictionary checks, and reproducible scripts are genuine strengths and raise the standard for computational-analytic work in this area. The paper is carefully scoped and does not overclaim.","major_comments":[],"minor_comments":[{"comment":"In Lemma 2.2, the decay O((1+|Re z|)^{-2}) is obtained by two Stieltjes integrations by parts; a one-sentence reminder that this is enough for absolute convergence via the Riemann–von Mangoldt local count N(t+1)−N(t)=O(log t) would make the appeal to the Guinand–Weil formula fully self-contained for readers outside the explicit-formula literature.","section":null},{"comment":"Corollary 2.7: the integral-domain argument for the Volterra convolution algebra of analytic germs is clean, but a brief parenthetical that the lowest-order coefficient is a nonzero beta-integral multiple would help readers who do not immediately recall the germ product formula.","section":null},{"comment":"Figure 1 caption and the worked example in §2.3: the numerical vector v is given to seven decimals; stating the exact rational form of (v2,v3,v4)=(1,0,−3)/√2 and the two linear conditions that fix (v0,v1) would make the example fully reproducible from the text alone.","section":null},{"comment":"Corollary 3.3(iii): the asymptotic B_T=(2N+1)ρ(log(T/2π)+1)/(π²T)(1+o(1)) is stated for fixed (c,N) as T→∞; a short remark that the o(1) absorbs both the h_+ expansion and the oscillatory integral after one integration by parts would clarify the error source.","section":null},{"comment":"Section 4: the three-route confirmation and the Arb LDL^T certificate at (c,N)=(100,200) are valuable; listing the precise working precision (bits) and the machine-readable artifact names in the text (in addition to the GitHub/Zenodo pointers) would further aid independent checking.","section":null},{"comment":"Notation: ρ=2π/L and Δ=L/(2π) are introduced early and used consistently, but a single display collecting L, Δ, ρ, a_T=T/ρ, and the even-sector embedding u would reduce cognitive load in §§2–3.","section":null}],"recommendation":"accept","confidential_remarks":"The manuscript is tightly scoped, mathematically careful, and unusually well-supported by released verification artifacts. It is a natural fit for a technical note or short article in analytic number theory / spectral approaches to zeta. I see no load-bearing gap; the softest external input (applicability of the Guinand–Weil formula to the class of Lemma 2.2) is standard and adequately checked. Accept is appropriate."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"This paper gives two clean finite theorems about the Connes–van Suijlekom / Connes–Consani–Moscovici Galerkin matrices for the truncated Weil form. First, every even coefficient vector v is transported by an explicit inverse-free chain (trig poly → Volterra kernel → compact Fourier weight → entire g_v) so that the cutoff-free quadratic value equals the sum of g_v over the nontrivial zeros exactly. Second, the omitted archimedean tail past any T > ρN is a rank-two Cauchy–Stieltjes Gram increment that is positive definite and strictly totally positive, yielding the two-sided budget B_T ∼ (2N+1)ρ log T / (π² T). Finite-T positivity certifies the cutoff-free sign; an eigenvalue below −B_T certifies a genuine negative; the inconclusive band is quantified.\n\nWhat is actually new is the closed-form transport, the exact 2N+1 source quotient, the non-collapsing pole-neutral family of dimension N−s−1, and especially the tail-order theorem with its elementary budget. The zero-sum identity itself is the classical Guinand–Weil formula applied to the induced admissible class; the paper says so in Remark 2.6 and does not oversell it. The proofs are elementary calculus plus classical Cauchy total positivity (Andréief + determinants). Entry identification with the CCM closed forms is checked, and the package ships Arb certificates, three-route dictionary checks, and partial sums over the first 512 zeros. The disclaimers (no RH, no prime-counting, no factoring) are explicit and correct.\n\nSoft spots are minor and proportional. The only external input is that the Volterra-induced g_v lies in the standard Guinand–Weil class (exponential type ≤ L, compact Fourier support, strip decay O((1+|Re z|)^{-2})); that is established by two Stieltjes integrations by parts and is the usual domain for the formula. The dictionary is one-way; no inverse is claimed. Total positivity is only for the isolated post-band archimedean increment. None of this undercuts the finite statements.\n\nThis is for people already working with Connes-style truncations or finite Weil positivity who need calibrated instruments rather than brute cutoffs. The math is sound, the artifacts are real, and the citation pattern is appropriate. I would send it to peer review and would bring it to reading group.","headline":"Exact finite dictionary and tail-order certification for Connes-style Weil truncations; solid elementary math with shipped verification, no RH claim.","tokens_in":16633,"tokens_out":601,"would_cite":true,"duration_ms":4909,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["11M26","11M06","15A42","47B36"],"pacs":[],"model":"grok-4.5","headline":"Every value of the truncated Weil form is an exact sum over zeta zeros, and finite cutoffs obey an explicit positivity budget.","keywords":["Weil quadratic form","Guinand–Weil explicit formula","Galerkin truncation","archimedean tail","total positivity","Riemann zeta zeros","finite certification"],"falsifier":"For a concrete even vector $v$ at fixed (c, N), recompute the closed-form matrix contraction $\\langle v, Q_\\infty v\\rangle$ and the partial zero-sum of $g_v$ over the first several hundred ordinates; if the residual fails to shrink consistently with the archimedean tail estimate (as already checked for the first 512 zeros), the dictionary identity fails.","tokens_in":16551,"feed_emoji":"∑","tokens_out":785,"duration_ms":6738,"temperature":0.7,"texified_at":"2026-08-05T21:12:52.383283+00:00","pith_summary":"This paper turns the finite Galerkin truncations of the Weil quadratic form into a calibrated instrument. First, every real even coefficient vector determines, by a closed chain of maps, a band-limited Guinand–Weil test function whose sum over the nontrivial zeros of zeta equals the quadratic value of the cutoff-free truncated matrix exactly. No limit in the band size and no numerical quadrature is required: the matrix entries themselves are exact zero sums. Second, the archimedean integral that must be cut off in any numerical computation has a totally positive tail past the Galerkin band. That tail order supplies a two-sided certification rule with an explicit budget that decays only like log T over T: finite-cutoff positivity certifies true positivity, a deep enough negative certifies a true negative, and a mild negative inside the budget band certifies nothing. Deep spectral scales that would demand astronomically large cutoffs become reachable by assembling the cutoff-free closed forms instead. The paper makes no claim about the Riemann hypothesis, primes, or factoring; it isolates what is already exact at every finite level and what finite numerics can and cannot certify about it.","texify_model":"deepseek-v4-flash","texify_usage":{"total_tokens":2944,"prompt_tokens":683,"completion_tokens":2261,"prompt_tokens_details":{"cached_tokens":0},"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":683,"completion_tokens_details":{"reasoning_tokens":1655}},"feed_headline":"Truncated Weil form values are exact sums over zeta zeros","feed_subtitle":"A closed dictionary and a log-T/T budget decide what finite cutoffs can certify","key_machinery":"The finite Guinand–Weil dictionary: the closed chain $v \\to T_v \\to K_v \\to \\hat{g}_v \\to g_v$ that sends a coefficient vector to an admissible band-limited test function, together with the rank-two density representation of the archimedean tail that proves total positivity and supplies the budget $B_T$.","core_discovery":"For every real even Galerkin vector $v$ the cutoff-free truncated Weil matrix evaluates the sum of an explicitly constructed band-limited test function $g_v$ over the nontrivial zeros of $\\zeta$ with multiplicity. Independently, the omitted archimedean tail past any cutoff $T$ larger than the Galerkin band is a strictly totally positive Cauchy–Stieltjes increment, so the finite-$T$ eigenvalues sandwich the true eigenvalues within an explicit budget $B_T$ that behaves like $\\frac{(2N+1)\\rho\\log T}{\\pi^2 T}$.","pith_inferences":[],"forward_implications":[],"fun_headline_variants":["Finite Guinand-Weil dictionary equates truncated forms to zeta zero sums","Galerkin vectors map exactly to band-limited sums over zeta zeros","Archimedean tail order yields log-T/T budget for Weil positivity","Closed dictionary makes every truncated Weil value an exact zero sum","Two-sided certification: finite cutoffs bound cutoff-free Weil eigenvalues"],"cache_read_input_tokens":128,"weakest_assumption_plain":"The induced test functions must lie in the classical class for which the Guinand–Weil explicit formula is known to hold with an absolutely convergent zero sum; the paper treats that membership as standard rather than re-proving the formula.","fun_headline_variants_meta":{"raw":{"variants":["Finite Guinand-Weil dictionary equates truncated forms to zeta zero sums","Galerkin vectors map exactly to band-limited sums over zeta zeros","Archimedean tail order yields log-T/T budget for Weil positivity","Closed dictionary makes every truncated Weil value an exact zero sum","Two-sided certification: finite cutoffs bound cutoff-free Weil eigenvalues"]},"model":"grok-4.5","effort":"low","cost_usd":0.003682,"raw_usage":{"total_tokens":1303,"prompt_tokens":938,"num_sources_used":0,"completion_tokens":96,"cost_in_usd_ticks":36820000,"prompt_tokens_details":{"text_tokens":938,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":269,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":938,"tokens_out":96,"duration_ms":2732,"temperature":1.0,"reasoning_tokens":269,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-12T06:46:28.004201+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"For a concrete even vector $v$ at fixed (c, N), recompute the closed-form matrix contraction $\\langle v, Q_\\infty v\\rangle$ and the partial zero-sum of $g_v$ over the first several hundred ordinates; if the residual fails to shrink consistently with the archimedean tail estimate (as already checked for the first 512 zeros), the dictionary identity fails.","supporting_citations":[],"review_version":1}