{"id":"6eecc3ec-9558-4e2d-959e-e9e4001e90bc","arxiv_id":"2507.02522","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"An exact spectral summation formula is proven for products of four Fourier coefficients of half-integral weight cusp forms; the arithmetic side is a sum of generalized class numbers of pairs of quadratic forms.","lead":"A new formula links weighted sums of products of four Fourier coefficients of half-integral weight cusp forms in Kohnen's subspace to generalized class numbers of pairs of quadratic forms. The identity extends the Bruggeman-Kuznetsov framework to fourth moments and opens a route to moment and distribution questions for modular forms.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The main theorem depends on the exact normalization of the Katok–Sarnak formula imported in Lemma 3.5(iii); an off-by-constant there would break the spectral identity.","rationale":"The paper's structure is sound: Lemma 2.2 gives the elementary side, Lemma 2.6 and Lemma 3.5 give the spectral side, and the test-function choices in Section 4 convert the spectral expressions into the stated Tχ-transform via standard hypergeometric identities. I checked the dimension of (3.14): with b_j(n) scaling like (4π|n|)^{-1/4}, the RHS scales like |δ|^{1/2}, matching the number of Heegner points, so no obvious dimensional inconsistency exists. The main residual risk is the imported Katok–Sarnak formula, whose normalization is acknowledged but not fully verified in the paper. Since the author explicitly flags the 4π-normalization conversion and the rest of the proof is detailed and coherent, I do not see a definite error. The reader's ACCEPT verdict stands, though an independent check of (3.14) would raise confidence from moderate to high.","tokens_in":29824,"tokens_out":22318,"duration_ms":236571,"concrete_test":"Independently re-derive (3.14) from the stated [I-L-T, Theorem 1.4] and [D-I-T, Proposition 6] with explicit normalizations: write u(z) with W_{0,it}(4π|m|y)e(mx), F_j(z) with W_{1/4 sgn(m),it}(4π|m|y)e(mx), and the Petersson norms on F_1 and F_4. If the derived identity is exactly (3.14), the concern is settled. If the constant differs, recompute the spectral terms (4.6)–(4.8) with the corrected constant; the discrepancy would show up as a uniform scaling of the cuspidal and Eisenstein contributions in Theorem 1.1.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Theorem 1.1 is assembled from the spectral decomposition of the inner product (4.1) and the Heegner-point evaluation Lemma 3.5. The decisive bridge is (3.14): (1/(u,u)_1) Σ_{Q∈Λ_δ} ω_D(Q)/M_Q u(z_Q) = 12|δ|^{3/4} b_j(D)b_j(δ/D). This is imported from [I-L-T, Theorem 1.4] and [D-I-T, Proposition 6], with the note that b_j(n) corresponds to b_ψ(n)(4π|n|)^{-1/4}. The constant 12 and the exact arguments D and δ/D depend on three normalizations simultaneously: the Whittaker normalization in (1.2), the Petersson inner products (·,·)_1 and (·,·)_4, and the Hecke normalization a(1)=1 of the Shimura lift. The paper does not display the full statement of [I-L-T, Theorem 1.4] or carry out the conversion term-by-term, so an undetected factor of 2, π, or |D|^{±1/2} in (3.14) would propagate linearly into the spectral side of Theorem 1.1. This is the same weak point the reader identified; it is real but not demonstrated to be wrong.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves an exact spectral summation formula (Theorem 1.1) for a fourth moment of Fourier coefficients of Maass cusp forms in Kohnen's subspace of weight 1/2, weighted by the Petersson norm of the Shimura lift, plus an Eisenstein contribution and a constant term, and equates this to generalized class numbers of pairs of integral binary quadratic forms. The proof computes the inner product of two automorphic functions M_{t,n,D,m} in two ways: by an elementary geometric unfolding (Lemma 2.2) into class numbers, and spectrally through the spectral theorem (4.1), Lemma 2.6, and a Heegner-point evaluation (Lemma 3.5). A carefully engineered pair of test functions then converts the spectral side into the stated Jacobi-type transform T_χ. A separate theorem, Theorem 1.2, gives an analogous integral representation for products of two negative Fourier coefficients. All infinite sums and integrals in Theorem 1.1 are shown to converge absolutely.","tokens_in":30028,"tokens_out":9506,"duration_ms":113932,"significance":"If the result is correct, it is a substantial new structural theorem: an exact fourth-moment formula for half-integral weight Maass forms in Kohnen's subspace, tied to arithmetic class numbers of pairs of quadratic forms. It goes beyond the classical two-coefficient Kuznetsov/Katok-Sarnak framework and complements the Blomer-Corbett fourth-moment results, while the method is a natural extension of the author's earlier triple-product work. The manuscript is careful and largely self-contained: convergence is controlled by explicit lemmas, the spectral decomposition is stated cleanly, and there are no fitted free parameters. The main caveat is that two decisive Heegner-point identities are imported from [I-L-T] and [D-I-T] with a normalization conversion that is asserted rather than demonstrated; this is a genuine correctness-risk even though no concrete error was found.","major_comments":[{"comment":"The constants in the main theorem depend crucially on the normalization conversion in Lemma 3.5(iii). Equation (3.14) is imported from [I-L-T, Theorem 1.4] after the one-sentence remark that b_j(n) corresponds to b_ψ(n)(4π|n|)^{-1/4}, together with a comment about positive versus negative definite classes; equation (5.19) is imported from [D-I-T, Proposition 6] and [I-L-T, Theorem 1.4] in the same way. Because Theorem 1.1 is homogeneous of degree one in the product of the four Fourier coefficients, any missing factor of 2, π, or |D|^{±1/2} in (3.14) would propagate linearly into the final identity, and the same applies to the factor 12√π in (5.19). I did not find an actual error, but this is a load-bearing step: the paper should restate the imported theorem in full in the present notation, identify the relevant eigenfunction ψ in [I-L-T, (1.14)] using Lemma 3.2(ii), and carry out the conversion term by term, or provide an appendix tabulating the Whittaker normalization in (1.2), the Petersson norms (·,·)_1 and (·,·)_4, the Hecke normalization a(1)=1, and the treatment of positive versus negative definite classes.","section":"§3.3, Eq. (3.14); §5, Eq. (5.19)"}],"minor_comments":[{"comment":"The proof of the hypergeometric identity (4.24)=(4.25) relies on [A-A-R, Corollary 3.3.5] and [S, (4.3.4.2)] at a crucial point; since these transformations have parameter and convergence hypotheses that are easy to misapply, please quote the exact identities or add a direct coefficient comparison.","section":"§4.1, Lemma 4.1"},{"comment":"The shorthand Γ(C−1/4±iz) is used where a product of two gamma factors is intended; this should be defined explicitly to avoid ambiguity.","section":"§4.1, Eqs. (4.13), (4.15), (4.22)"},{"comment":"There are several typographical errors that should be corrected, including 'fou r' in the abstract, 'Diriclet' in Section 1.3, 'rght-hand side' in the proof of Lemma 4.1, and 'nonpsitive' in the same proof.","section":"Throughout"},{"comment":"The multiplier system ν is invoked in the transformation formula for B0(z) before its domain of definition is fully described; a one-sentence definition or reference at that point would improve readability.","section":"§1.2"}],"recommendation":"major_revision","confidential_remarks":"This is a strong and original contribution. My recommendation of major revision is driven solely by the need to verify, in detail, the normalization bridge between the paper's Fourier coefficients and the Katok-Sarnak formula imported from [I-L-T] and [D-I-T]. If the author can provide that verification, or point to exact equations with parameter matching, I would support acceptance. I do not think new numerical experiments or new theorems are required."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The one thing to know: this is a real result, not a repackaging. Bíró proves a spectral summation formula for products of four Fourier coefficients of half-integral weight cusp forms in Kohnen’s subspace, with the arithmetic side given by generalized class numbers of pairs of quadratic forms. That goes cleanly beyond the earlier [B-C] special case, where two of the four factors were first coefficients. The theorem handles arbitrary negative discriminants and two positive fundamental discriminants, and the statement is sharp enough to be useful for moment estimates and distribution questions.\n\nThe paper earns its keep on execution. The proof is long but structured: an inner product of two automorphic functions of type M, an elementary evaluation in terms of class numbers, a spectral evaluation via the spectral theorem, and then a careful choice of test functions to extract the four-coefficient identity. Convergence is handled explicitly, including the absolute convergence of every sum and integral. Lemma 2.2 and Lemma 2.6 are proved in detail. The author also clearly states how the argument adapts to other sign patterns, and Theorem 1.2 gives the two-negative-coefficient case needed there. I see no circularity and no fitted parameters. The result is genuinely new.\n\nThe soft spots are real but not fatal. The load-bearing step is Lemma 3.5(iii), equation (3.14), which converts a Heegner-point sum into a product of two Fourier coefficients. As the stress-test note says, this is imported from [I-L-T] and [D-I-T] with a normalization conversion that is described in words rather than displayed term-by-term. If an off-by-constant factor were hiding in that conversion, it would propagate linearly into Theorem 1.1. I found no reason to think the constant is wrong, but the paper would be stronger if it showed the full statement of the imported theorem and the conversion explicitly. Lemma 4.1, the hypergeometric identity, is also not fully verified in the text; it looks like standard manipulation, but I would want a referee to check it. These are verification gaps, not demonstrated flaws.\n\nWho gets value: analytic number theorists working on half-integral weight moments, Shimura lifts, and spectral summation formulas. The paper deserves a serious referee, not a desk reject. My recommendation: send it out, and ask the referee to focus on the normalization matching in Lemma 3.5(iii) and the hypergeometric identity in Lemma 4.1. If those check out, the paper is solid.","headline":"A genuine four-coefficient spectral summation formula for weight-1/2 cusp forms, built on known machinery; the main risk is the imported Katok–Sarnak normalization, which deserves a careful referee check.","tokens_in":30555,"tokens_out":1703,"would_cite":true,"duration_ms":21936,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["11F37","11F72"],"pacs":[],"model":"deepseek-v4-flash","headline":"Products of four half-integral weight Fourier coefficients, summed over the spectrum, equal generalized class numbers of pairs of quadratic forms.","keywords":["Shimura lift","Kohnen subspace","half-integral weight","Maass cusp forms","pairs of quadratic forms","generalized class numbers","spectral summation formula","Zagier L-functions"],"falsifier":"Compute both sides of Theorem 1.1 numerically for a small explicit case, for example $\\delta_1=\\delta_2=-3$, $D_1=D_2=1$, with test function $\\chi(z)=e^{-z^2}$. The spectral side can be evaluated from the first few weight-$1/2$ Maass cusp forms in Kohnen's subspace, and the arithmetic side from the explicit class-number formulas for $h_{1,1}(-3,-3,f)$; any disagreement beyond numerical precision would refute the identity.","tokens_in":29582,"feed_emoji":"🔢","tokens_out":9013,"duration_ms":84092,"temperature":0.7,"pith_summary":"This paper proves an exact identity between two very different-looking objects. On one side is a spectral sum: products of four Fourier coefficients of half-integral weight cusp forms in Kohnen's subspace, weighted by the norms of their Shimura lifts and summed over the whole spectrum, plus a contribution from Eisenstein series and a constant term. On the other side is an arithmetic sum over $SL_2(\\mathbb{Z})$-equivalence classes of pairs of integral quadratic forms with prescribed discriminants and codiscriminant, weighted by genus characters. The identity is the natural fourth-moment analogue of the Bruggeman–Kuznetsov formula, which connects products of two coefficients to Kloosterman sums. A companion theorem (Theorem 1.2) gives an integral representation for products of two negative coefficients, which the paper uses to explain how the main identity extends to other sign patterns. Every sum and integral in the statement is absolutely convergent.","feed_headline":"Spectral sums of four weight-1/2 coefficients equal class-number sums","feed_subtitle":"Products of four Shimura-lifted coefficients matched to weighted counts of pairs of quadratic forms.","key_machinery":"The load-bearing mechanism is the Shimura lift together with a Katok–Sarnak-type formula. The Shimura lift sends a weight $1/2$ Maass cusp form $F_j$ in Kohnen's subspace to an even, Hecke-normalised Maass cusp form $\\mathrm{Shim}\\,F_j$ of weight $0$ for $SL_2(\\mathbb{Z})$, and the paper uses a generalized Kohnen–Zagier theorem to show this map is a bijection onto the even spectral side. The Katok–Sarnak formula (from [I-L-T] and [D-I-T]) then expresses a weighted sum of values of such a Maass form over Heegner points of discriminant $\\delta$ as essentially the product $b_j(D)b_j(\\delta/D)$, which is where the four Fourier coefficients arise. The spectral computation is performed on the functions $M_{t,n,D,m}(z)$, point-pair averages over conjugacy classes of matrices, whose inner product has an elementary expression in terms of the class numbers $h_{D_1,D_2}$; comparing the two evaluations yields the identity. The integral transform $T_\\chi$ is a special Jacobi transform, whose inversion is known, so the identity can be re-expressed with a general test function on the arithmetic side.","core_discovery":"Theorem 1.1 states that for integers $\\delta_1,\\delta_2<0$ and positive fundamental discriminants $D_1,D_2$ with $D_i\\mid\\delta_i$ and $\\delta_i/D_i\\equiv 0,1\\pmod 4$, the quantity\n$$144\\pi|\\delta_1\\delta_2|^{-3/4}\\sum_{j\\ge 1}(\\mathrm{Shim}\\,F_j,\\mathrm{Shim}\\,F_j)_1\\, b_j(D_1)b_j(\\delta_1/D_1)b_j(\\delta_2/D_2)b_j(D_2)\\chi(r_j)$$\nplus a constant term and an Eisenstein integral is exactly equal to\n$$E_{\\delta_1,\\delta_2,D_1,D_2}T_\\chi(0)+\\sum_{f\\in\\mathbb{Z},\\,$f^{2}$>|\\delta_1\\delta_2|} h_{D_1,D_2}(\\delta_1,\\delta_2,f)\\,T_\\chi\\!\\left(\\frac{$f^{2}$}{|\\delta_1\\delta_2|}-1\\right),$$\nwhere $h_{D_1,D_2}$ counts weighted $SL_2(\\mathbb{Z})$-classes of pairs of quadratic forms with codiscriminant $f$ and $E_{\\delta_1,\\delta_2,D_1,D_2}$ is a similar degenerate class sum. The spectral side is a weighted fourth moment of Fourier coefficients of weight $1/2$ Maass cusp forms in Kohnen's subspace; the arithmetic side is a sum of generalized class numbers transformed by a Jacobi-type integral transform $T_\\chi$. The paper proves the identity by computing one inner product of two automorphic point-pair functions in two ways, once by an elementary orbit count and once by the spectral theorem.","pith_inferences":["Beyond the paper: the same two-evaluation mechanism should produce fourth-moment identities with four positive or three positive coefficients, since the paper sketches the needed modifications in Section 1.8; making those formulas explicit is a natural next step.","Beyond the paper: because the class-number sum is transformed by an invertible integral operator, one could choose sequences of test functions that localize at large spectral parameter to extract asymptotic information on the generalized class numbers from the spectral side.","Beyond the paper: the Heegner-point identity in Lemma 3.5(iii) is a generalized Kohnen–Zagier relation; pairing it with known variance estimates for half-integral weight coefficients might yield new results on the distribution of Heegner points, though the paper does not pursue this."],"forward_implications":["The generalized class numbers $h_{D_1,D_2}(\\delta_1,\\delta_2,f)$ acquire a spectral interpretation as weighted fourth moments of half-integral weight Fourier coefficients, which is new for products of four coefficients.","Because $T_\\chi$ is an invertible Jacobi transform, the identity can be restated with a general test function on the arithmetic side, making the formula adaptable to different spectral cut-offs.","For $D_1=D_2=1$, the class numbers $h_{1,1}(\\delta_1,\\delta_2,f)$ have explicit elementary expressions, so in that case Theorem 1.1 is a completely explicit identity.","Theorem 1.2 gives an integral representation for products of two negative Fourier coefficients, and Section 1.8 explains how the same method yields analogous summation formulas for sign patterns other than two positive and two negative coefficients."],"supporting_citations":[{"why":"Supplies the Katok–Sarnak formula for higher weights (Theorem 1.4) that turns Heegner-point sums into products of four Fourier coefficients in Lemma 3.5(iii).","marker":"[I-L-T]"},{"why":"Supplies Proposition 6, the geometric identity used in (5.19), and the definition of the Shimura lift valid without the condition $b_j(1)\\neq 0$.","marker":"[D-I-T]"},{"why":"Lemma 2.2 expresses the inner product of two automorphic functions as a weighted sum of class numbers of pairs of quadratic forms, giving the arithmetic side of the formula.","marker":"[B2]"},{"why":"Lemma 2 computes the integral of $M_{t,n,D,m}$ against a Maass form as a Heegner-point sum, a key input for the spectral side.","marker":"[B1]"},{"why":"Defines Kohnen's subspace, the Hecke operators $T_{p^2}$, and the Shimura lift for weight $1/2$ Maass forms.","marker":"[K-S]"},{"why":"Theorem 1.2, a generalized Kohnen–Zagier formula, is used to show the map $j\\mapsto \\mathrm{Shim}\\,F_j$ is a bijection onto even Hecke-normalised Maass cusp forms.","marker":"[B-M]"},{"why":"Provides the polynomial bound on Fourier coefficients needed for absolute convergence of the spectral sum in Lemma 4.2.","marker":"[Du]"},{"why":"Theorem 15.5 supplies the spectral expansion used to compute the inner product by the spectral theorem.","marker":"[I-K]"}],"fun_headline_variants":["Weight-1/2 moments equal sums of class numbers","Fourth moments of half-integral weight forms hit class sums","Kohnen subspace: spectral sums become class number sums","Half-integral weight spectral identity for pairs of quadratic forms"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The identity rests on the Katok–Sarnak formula being exactly right with the paper's chosen normalization of Fourier coefficients, namely $b_j(n)$ matching $b_\\psi(n)(4\\pi|n|)^{-1/4}$ in [I-L-T]'s notation; a mismatch there breaks Lemma 3.5(iii) and Theorem 1.2.","fun_headline_variants_meta":{"raw":{"variants":["Weight-1/2 moments equal sums of class numbers","Fourth moments of half-integral weight forms hit class sums","Kohnen subspace: spectral sums become class number sums","Half-integral weight spectral identity for pairs of quadratic forms"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000393,"raw_usage":{"total_tokens":2064,"prompt_tokens":943,"completion_tokens":1121,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":559,"completion_tokens_details":{"reasoning_tokens":1054}},"tokens_in":559,"tokens_out":1121,"duration_ms":11497,"temperature":1.0,"reasoning_tokens":1054,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T20:27:36.073620+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute both sides of Theorem 1.1 numerically for a small explicit case, for example $\\delta_1=\\delta_2=-3$, $D_1=D_2=1$, with test function $\\chi(z)=e^{-z^2}$. The spectral side can be evaluated from the first few weight-$1/2$ Maass cusp forms in Kohnen's subspace, and the arithmetic side from the explicit class-number formulas for $h_{1,1}(-3,-3,f)$; any disagreement beyond numerical precision would refute the identity.","supporting_citations":[],"review_version":1}