{"id":"61900627-0101-4b18-982e-27ee5c5bbb7a","arxiv_id":"2504.12604","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"MacWilliams weight enumerator identities for linear codes over Z_k are re-derived via theta functions, extended to arbitrary genus, and linked to cyclotomic-field theta functions for prime k.","lead":"This paper proves identities between weight enumerators of linear codes over the modular ring Z_k and theta functions of associated lattices, and extends a known higher-genus identity to all k. The value is in connecting coding theory invariants with lattice and modular form theory, though much of the content extends or re-proves known results.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"A_j=A_{k-j} (and ϑ_j=ϑ_{p-j}) make the theta-map non-Zariski-dense, so Theorem 2 cannot yield the complete weight-enumerator identity; the algebraic independence sought in Section VI is impossible.","rationale":"The reader identified algebraic independence of A_0,...,A_{k−1} as the weakest assumption, and I agree; the present review strengthens that concern by showing the independence is not merely missing but impossible because A_j=A_{k−j}. The main theorems are not all undermined: Theorem 3 is proved directly by Fourier transform and Poisson summation, and it does prove the complete weight enumerator MacWilliams identity in genus g for all k; Theorem 1 and Theorem 4 are correct functional identities but, because of the symmetry, they express only a symmetrized weight enumerator, not the complete one. The central overclaim is therefore in the abstract, the introduction, and the conclusion's claim that Theorem 2 provides a theta-function route to the complete weight enumerator identity. The correctable fix is to restate Theorem 2 and Theorem 4 as symmetrized identities, to cite Theorem 3 (with g=1) as the proof of the complete identity, and to remove or correct the Section VI open problem about algebraic independence. The reader's CONDITIONAL verdict already captures the need for such corrections, so no verdict change is required. Minor additional issue: the proof of Theorem 2 cites Proposition 2 where Proposition 3 is intended, but this is a typographical error rather than a mathematical gap.","tokens_in":19915,"tokens_out":22157,"duration_ms":243572,"concrete_test":"Analytical check: in Definition 7, write the sums for A_j and A_{k−j}; the change of summation variable x↦−x maps kZ+j onto kZ+(k−j) term-by-term, so A_j(z)−A_{k−j}(z)=0 as a formal q-series. For k=3 this is visible in the q-expansions with q=e^{π i z/3}: A_1=q+q^4+q^16+⋯ and A_2=q+q^4+q^16+⋯. A parallel reindexing in Section V gives ϑ_j−ϑ_{p−j}=0. These equalities are enough to disprove the Zariski-density premise; a positive control is to verify that the symmetrized identity (Corollary 2) holds at the point (A_0,…,A_{k−1}), while a polynomial identity in variables with X_j≠X_{k−j} is not reachable from that point.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The decisive weakness is not merely that algebraic independence is unproved; it is false. In Definition 7, A_j(z)=Σ_{x∈kZ+j} e^{π i z x^2/k}. The substitution x↦−x is a bijection from kZ+j to kZ+(k−j), and the exponent depends on x^2; hence A_j=A_{k−j} identically for every j. Thus (A_0,...,A_{k−1}) lies in the proper subspace {X_j=X_{k−j}}; for k≥3 the map z↦(A_0(z),...,A_{k−1}(z)) is not Zariski dense in C^k. Consequently, the equality obtained in the proof of Theorem 2 (Eqs. (3.2)-(3.3)) is only a functional identity on that subspace. Without Zariski density, one cannot replace A_j by independent variables X_j and conclude the polynomial complete weight enumerator MacWilliams identity; at best one gets the symmetrized identity of Corollary 2. The conclusion's request for algebraic independence of A_0,...,A_{k−1} therefore cannot succeed. The same symmetry afflicts Theorem 4: ϑ_j=ϑ_{p−j} under x↦−x, so the theta function of Γ_C sees only the symmetrized (Lee-type) enumerator, not the complete Hamming weight enumerator. This does not make Theorem 3 false—its proof by Fourier/Poisson summation is independent of the theta substitution—but it invalidates the claim that the complete weight enumerator MacWilliams identity is obtained from the one-variable theta functions in Theorem 2.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies linear codes over the finite ring Z_k via lattices and theta functions. It formulates k auxiliary theta functions A_j(z), proves that the theta function of the lattice associated to a code C equals the complete weight enumerator of C evaluated at the A_j (Theorem 1), and then derives a functional identity for W_{C^⊥} evaluated at the same functions (Theorem 2). From this it obtains the symmetrized weight enumerator MacWilliams identity (Corollary 2). It further proves the genus-g complete weight enumerator MacWilliams identity for all positive k by Fourier transform and Poisson summation (Theorem 3), and for odd primes p gives an identity between the theta function of a lattice over a cyclotomic field and the complete weight enumerator of a code C ⊂ F_p^n with C ⊂ C^⊥ (Theorem 4). The paper's framing claims that these results extend the Bannai-Dougherty-Harada-Oura and van der Geer-Hirzebruch identities from even k to all k and from Lee to complete weight enumerators, respectively.","tokens_in":20242,"tokens_out":11088,"duration_ms":110352,"significance":"If all claims were valid, the paper would provide a theta-function interpretation of the MacWilliams identity for codes over Z_k for arbitrary k, a genus-g extension for all k, and a complete-weight-enumerator analogue of the Alpbach theorem. The paper contains some correct and useful pieces: Theorem 3 is a genuinely correct, self-contained proof of the genus-g complete weight enumerator MacWilliams identity for every positive integer k, extending the even-k case of Bannai et al.; Theorem 1 is a correct lattice-theta-function factorization; and the functional identity in Theorem 2 is correctly derived by Poisson summation. However, the central claimed implication from Theorem 2 to the polynomial complete weight enumerator MacWilliams identity fails because the required algebraic independence is not only unproved but is in fact false: the functions A_j satisfy A_j = A_{k-j}, so the theta map is never Zariski dense. Similarly, in the cyclotomic setting the functions ϑ_j satisfy ϑ_j = ϑ_{p-j}, so Theorem 4 does not deliver the advertised complete Hamming weight content.","major_comments":[{"comment":"For every j, A_j(z) = A_{k-j}(z) identically, because x ↦ -x is a bijection between kZ+j and kZ+(k-j) and the exponent in Definition 7 depends only on x^2. Hence the map z ↦ (A_0(z),...,A_{k-1}(z)) takes values in the proper closed subspace {X_j = X_{k-j}} of C^k, and for k ≥ 3 it is not Zariski dense in C^k. The authors state in Section VI that one must prove algebraic independence of A_0,...,A_{k-1} to show that Theorem 2 is equivalent to the complete weight enumerators MacWilliams identity; that algebraic independence is not merely unproved but impossible. Consequently the proof of Theorem 2 establishes only a functional identity on this symmetric subspace, and the only polynomial identity that can be concluded is the symmetrized identity of Corollary 2. The abstract's assertion that the complete weight enumerators MacWilliams identity is obtained from theta functions is therefore not supported by the proof.","section":"Section III, Definition 7 and Eqs. (3.2)-(3.3)"},{"comment":"The same symmetry invalidates the advertised interpretation of Theorem 4. The ideal B is stable under negation, and Tr_{K^+/Q}(x x̄/p) is invariant under x ↦ -x, so ϑ_j(z) = ϑ_{p-j}(z) identically for every j. The identity ϑ_{Γ_C}(z) = W_C(ϑ_0(z),...,ϑ_{p-1}(z)) therefore reduces to the symmetrized Lee-type identity S_C(ϑ_0,...,ϑ_{(p-1)/2}), with the variables corresponding to the pairs {j,p-j} being redundant. Thus Theorem 4 is not a generalization of the van der Geer-Hirzebruch Alpbach theorem from Lee weight to complete Hamming weight; it is the Lee-weight identity written with duplicated variables. The same objection applies to the higher-dimensional version in Eq. (5.1).","section":"Section V, Theorem 4 and Eq. (5.1)"},{"comment":"The proof of Theorem 3 by finite Fourier transform and Poisson summation is correct and is independent of the theta-function machinery. This theorem, giving the genus-g complete weight enumerator MacWilliams identity for all positive integers k, is the paper's principal sound result. However, because it is proved directly and does not use the A_j or ϑ_j functions, it does not rescue the paper's central claim of a theta-function derivation of the complete weight enumerator MacWilliams identity. A revised version would need to reframe the contribution around Theorem 3 and the symmetrized corollaries rather than around the invalid theta-function derivation.","section":"Section IV, Theorem 3"}],"minor_comments":[{"comment":"There are typographical errors: 'Hammi ng' in the abstract and 'fould' in the Conclusion should be 'Hamming' and 'found', respectively.","section":"Abstract and Section VI"},{"comment":"The notation w_a(y_j) in the Fourier transform computation is introduced rather abruptly; it would be clearer to define it explicitly as the indicator that the column y_j equals a, i.e., w_a(y_j) = 1 if y_j = a and 0 otherwise.","section":"Section IV, proof of Theorem 3"},{"comment":"In the factorization of the sum over ρ^{-1}(c), the notation ϑ_{c_i}(z) uses elements c_i ∈ F_p as indices; writing ϑ_{c_i mod p}(z) or ϑ_{[c_i]}(z) would avoid confusion with the integer representatives.","section":"Section V, proof of Theorem 4"}],"recommendation":"reject","confidential_remarks":"The algebraic-independence objection is decisive for the paper's stated purpose: the needed independence is not just missing but is false, so the manuscript's central claim that it obtains the complete weight enumerator MacWilliams identity from theta functions cannot be repaired within the current scope. The remaining correct content, especially Theorem 3, is a modest extension of known finite-Fourier arguments and does not by itself meet the novelty bar suggested by the paper's framing. I would encourage the authors to consider a separate, more carefully scoped submission centered on Theorem 3 and the symmetrized identities."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper has two genuinely correct pieces: the genus-g complete weight enumerator MacWilliams identity for all positive k (Theorem 3), proved cleanly by Fourier transform and Poisson summation, and the cyclotomic lattice theta identity (Theorem 4), which is also a direct computation. Theorem 3 is a legitimate extension of Bannai et al.'s even-k result to arbitrary k, though it is routine rather than deep. Theorem 1 is standard construction-A material, correctly stated. The citation pattern is fine, and the self-citations are contextual; the proofs do not lean on them.\n\nThe main advertised result does not survive contact with the paper. The functions A_j defined in Definition 7 satisfy A_j = A_{k-j} identically, because x \\mapsto -x is a bijection from kZ+j to kZ+(k-j) and the exponent depends on x^2. So the vector (A_0(z),...,A_{k-1}(z)) always lies in the proper subspace X_j = X_{k-j}; it is not Zariski dense in C^k. That means the equality proved in Theorem 2 is only a functional identity on that subspace. You cannot replace the A_j by independent polynomial variables and conclude the polynomial complete weight enumerator MacWilliams identity. The authors themselves flag the need for algebraic independence in Section VI, but that independence is not merely unproved; it is false. The same symmetry afflicts Theorem 4: the theta functions satisfy \\vartheta_j = \\vartheta_{p-j}, so the theta function of the associated lattice sees only the symmetrized (Lee-type) enumerator, not the full Hamming complete weight enumerator. Calling Theorem 4 a generalization of van der Geer–Hirzebruch to complete weight enumerators overstates what the identity delivers.\n\nThere is also a minor mechanical slip: the proof of Theorem 2 invokes Proposition 2 where it needs Proposition 3.\n\nBottom line: the paper is not a throwaway. Theorem 3 is a correct, citable result, and the failure mode of the theta-function approach is instructive. But the framing in the abstract and introduction is wrong, and the main claimed interpretation cannot be repaired within the current method. A serious referee should be sent this, with the expectation of major revision: keep Theorem 3 as the central contribution, state the symmetrized nature of the theta-function identities explicitly, and remove the claim that the full MacWilliams identity has been derived from theta functions.","headline":"Correct but modest genus-g MacWilliams result sits inside a paper whose headline claim—deriving the full complete weight enumerator MacWilliams identity from one-variable theta functions—is invalid, because the theta functions are symmetric and the algebraic independence the authors say they need is impossible.","tokens_in":20787,"tokens_out":2621,"would_cite":false,"duration_ms":29176,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["94B05","11F27","11T71"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that the complete weight enumerator MacWilliams identity for linear codes over the finite ring $\\mathbb{Z}_k$ holds for every positive integer $k$, extends the genus-$g$ identity beyond the even-$k$ case, and, for odd…","keywords":["linear codes over Z_k","MacWilliams identity","complete weight enumerators","theta functions","associated lattices","cyclotomic fields","genus g MacWilliams identity","Poisson summation formula"],"falsifier":"For a fixed small $k$ (for example $k=2$ or $k=3$), compute the $q$-expansions of $A_0,\\ldots,A_{k-1}$ to sufficiently high order and use linear algebra to look for a nonzero homogeneous polynomial relation among them. Finding such a relation would show that Theorem 2 does not by itself yield the polynomial complete weight enumerator MacWilliams identity; proving that no such relation exists, for instance by showing the image of $z\\mapsto(A_0(z),\\ldots,A_{k-1}(z))$ is Zariski dense, would complete the paper's stated open step.","tokens_in":19685,"feed_emoji":"🔢","tokens_out":16773,"duration_ms":152845,"temperature":0.7,"pith_summary":"The paper's aim is to settle the MacWilliams identity for codes over $\\mathbb{Z}_k$ for every positive integer $k$ by viewing a code through its associated lattice. It shows that the $\\theta$ function of the lattice $\\Gamma_C = \\frac{1}{\\sqrt{k}}\\rho^{-1}(C)$ equals the complete weight enumerator of $C$ evaluated at $k$ auxiliary $\\theta$ functions, and uses Poisson summation to turn this into complete and symmetrized weight enumerator identities. The genus-$g$ identity, previously proved only for even $k$, is stated and proved for all $k$ via a Fourier transform over $\\mathbb{Z}_k^g$. For odd primes $p$, the paper proves that the $\\theta$ function of the cyclotomic lattice attached to a self-orthogonal code $C \\subset \\mathbb{F}_p^n$ equals the Hamming-weight complete weight enumerator $W_C(\\vartheta_0,\\ldots,\\vartheta_{p-1})$, generalizing the Lee-weight $\\theta$ identity over cyclotomic fields. The paper leaves one step open: converting the $\\theta$-function evaluation in Theorem 2 into the full polynomial identity requires algebraic independence of the auxiliary $\\theta$ functions, which it says it has not yet proved.","feed_headline":"Theta functions prove MacWilliams identity over every ring Z_k","feed_subtitle":"A genus-g identity that once required even k now covers all k, and prime codes get a cyclotomic theta version.","key_machinery":"The load-bearing objects are the $k$ $\\theta$ functions $A_j(z)=\\sum_{x\\in k\\mathbb{Z}+j} e^{\\pi i z x^2/k}$ for $j=0,\\ldots,k-1$, together with the associated lattice $\\Gamma_C=\\frac{1}{\\sqrt{k}}\\rho^{-1}(C)$. Their Poisson summation transform, $A_j(-1/z)=\\frac{1}{\\sqrt{k}}(z/i)^{1/2}\\sum_{m=0}^{k-1} e^{2\\pi i j m/k}A_m(z)$, is exactly the Fourier matrix that appears in the MacWilliams identity, so the $\\theta$ function of the associated lattice becomes a weighted sum of the $A_j$'s. In genus $g$, the same Fourier kernel appears as $T=(\\eta^{a\\cdot b})_{a,b\\in\\mathbb{Z}_k^g}$, and the Poisson summation formula over $\\mathbb{Z}_k^{g\\times n}$ carries the proof. For $k=p$ prime, the machinery moves to the cyclotomic field $K=\\mathbb{Q}(\\xi)$, $\\xi=e^{2\\pi i/p}$, with $\\theta$ functions $\\vartheta_j(z)=\\sum_{x\\in\\mathfrak{B}+j} e^{2\\pi i z \\mathrm{Tr}_{K^+/\\mathbb{Q}}(x\\bar{x}/p)}$ over the principal ideal $\\mathfrak{B}=\\langle1-\\xi\\rangle$; the same factorization argument gives the complete weight enumerator identity.","core_discovery":"The central discovery is a dictionary between codes and lattices that makes the MacWilliams identity a $\\theta$-function statement. For any linear code $C \\subset \\mathbb{Z}_k^n$, with associated lattice $\\Gamma_C = \\frac{1}{\\sqrt{k}}\\rho^{-1}(C)$, the paper proves $\\vartheta_{\\Gamma_C}(z) = W_C(A_0(z),\\ldots,A_{k-1}(z))$, where the $A_j$ are the $\\theta$ series attached to the residue classes of $\\mathbb{Z}$ modulo $k$ (Theorem 1). This is the starting point for Theorem 2, the genus-1 identity evaluated at the $A_j$, and for the symmetrized identity. The unrestricted claim is Theorem 3: for every positive integer $k$, the genus-$g$ complete weight enumerator satisfies $C_{C^\\perp,g}(z_a) = \\frac{1}{|C|^g} T\\,C_{C,g}(z_a)$ with $T=(\\eta^{a\\cdot b})$, $\\eta=e^{2\\pi i/k}$, proved by Fourier transform and Poisson summation. For odd prime $p$, Theorem 4 gives $\\vartheta_{\\Gamma_C}(z)=W_C(\\vartheta_0(z),\\ldots,\\vartheta_{p-1}(z))$ for $C\\subset C^\\perp$, with $\\Gamma_C=\\rho^{-1}(C)\\subset D^n$ in the cyclotomic integer ring and with the $\\vartheta_j$ defined by the ideal $\\mathfrak{B}=\\langle1-\\xi\\rangle$. The paper states that Theorem 2 is equivalent to the ordinary complete weight enumerator MacWilliams identity only if the functions $A_0,\\ldots,A_{k-1}$ are algebraically independent, and it lists that proof as an open problem.","pith_inferences":["The open algebraic-independence step is not needed to obtain the MacWilliams identity itself, since Theorem 3 with $g=1$ already proves the polynomial identity by Fourier analysis; the open step affects only the claim that the theta-function route recovers it.","The cyclotomic construction for primes $p$ suggests a testable extension to composite $k$ through the integers of an appropriate cyclotomic field, which the authors flag as future work; the main obstruction to check is whether the trace form remains positive definite.","If the $A_j$ turn out to be algebraically dependent, the theta-function proof could still be rescued by replacing the evaluation point with a generic point of the resulting algebraic variety, though that would require a different modular argument.","Because the matrix $T$ in Theorem 3 is the character table of the finite abelian group $\\mathbb{Z}_k^g$, the identity is a Fourier-duality statement about weight distributions, which may connect the result to finite abelian harmonic analysis beyond coding theory."],"forward_implications":["For every positive integer $k$, the genus-$g$ complete weight enumerator MacWilliams identity holds for codes over $\\mathbb{Z}_k$, removing the even-$k$ condition of the earlier genus-$g$ result.","Taking $g=1$ in Theorem 3 gives the ordinary complete weight enumerator MacWilliams identity for all $k$, and the symmetrized weight enumerator version follows by identifying $a$ with $-a$.","For a self-dual code over $\\mathbb{Z}_k$, the complete weight enumerator is invariant under the rotation described in Corollary 1.","For odd primes $p$, a self-orthogonal code over $\\mathbb{F}_p$ gives a cyclotomic lattice whose theta function equals its Hamming complete weight enumerator, and the same identity holds for $z\\in\\mathbb{H}^{p-1}$.","If the missing algebraic independence is established, Theorem 2 becomes a genuine modular-form proof of the usual MacWilliams identity."],"supporting_citations":[{"why":"Gives the genus-$g$ complete weight enumerator MacWilliams identity for even $k$ that Theorem 3 extends to all $k$, and the Type II code/lattice correspondence used in Section II.","marker":"[2]"},{"why":"Supplies the standard Construction A dictionary and the Lee-weight theta identity over cyclotomic fields that Theorem 4 generalizes.","marker":"[12]"},{"why":"Original source of the Lee-weight theta identity for codes over $\\mathbb{F}_p$ attached to cyclotomic lattices.","marker":"[18]"},{"why":"Original MacWilliams distribution theorem over finite fields that the paper generalizes to codes over $\\mathbb{Z}_k$.","marker":"[27]"},{"why":"First complete weight enumerator MacWilliams identity, the genus-1 statement the paper re-derives through theta functions.","marker":"[28]"},{"why":"Model for the Fourier-transform and Poisson-summation proof used in Theorem 3, developed for Galois rings.","marker":"[39]"},{"why":"Authors' earlier complete weight enumerator MacWilliams identity for $\\mathbb{Z}_k$ that the theta-function Theorem 2 reformulates.","marker":"[47]"}],"fun_headline_variants":[],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the $k$ $\\theta$ functions $A_0(z),\\ldots,A_{k-1}(z)$ are algebraically independent over $\\mathbb{C}$ — that is, no nonzero polynomial in $k$ variables vanishes after the substitution — so an identity that holds at these special functions is an identity of polynomials; the paper says this proof is still missing in its conclusion.","fun_headline_variants_meta":{"error":"DeepSeek 429: {\"error\":{\"message\":\"Too many requests. Your current concurrency is 128, which exceeds your concurrency limit of 117 based on your remaining balance. Please top up your balance to restore your concurrency.\",\"type\":\"rate_limit_error\",\"param\":null,\"code\":\"invalid_request_error\"}}"},"cache_creation_input_tokens":0},"created_at":"2026-08-16T12:29:47.464759+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"For a fixed small $k$ (for example $k=2$ or $k=3$), compute the $q$-expansions of $A_0,\\ldots,A_{k-1}$ to sufficiently high order and use linear algebra to look for a nonzero homogeneous polynomial relation among them. Finding such a relation would show that Theorem 2 does not by itself yield the polynomial complete weight enumerator MacWilliams identity; proving that no such relation exists, for instance by showing the image of $z\\mapsto(A_0(z),\\ldots,A_{k-1}(z))$ is Zariski dense, would complete the paper's stated open step.","supporting_citations":[{"cited_title":"Type II co des, even unimodular lattices, and invariant rings,","cited_arxiv_id":null,"evidence_quote":"Gives the genus-$g$ complete weight enumerator MacWilliams identity for even $k$ that Theorem 3 extends to all $k$, and the Type II code/lattice correspondence used in Section II."},{"cited_title":"Lattices and codes: A course partially bas ed on lectures by Friedrich Hirzebruch,","cited_arxiv_id":null,"evidence_quote":"Supplies the standard Construction A dictionary and the Lee-weight theta identity over cyclotomic fields that Theorem 4 generalizes."},{"cited_title":"Gesammelte Abhandlungen,","cited_arxiv_id":null,"evidence_quote":"Original source of the Lee-weight theta identity for codes over $\\mathbb{F}_p$ attached to cyclotomic lattices."},{"cited_title":"A theorem on the distribution of wei ghts in a systematic code,","cited_arxiv_id":null,"evidence_quote":"Original MacWilliams distribution theorem over finite fields that the paper generalizes to codes over $\\mathbb{Z}_k$."},{"cited_title":"The MacWi lliams identities for nonlinear codes,","cited_arxiv_id":null,"evidence_quote":"First complete weight enumerator MacWilliams identity, the genus-1 statement the paper re-derives through theta functions."},{"cited_title":"The MacWilliams identity for linear codes over Galois rings,","cited_arxiv_id":null,"evidence_quote":"Model for the Fourier-transform and Poisson-summation proof used in Theorem 3, developed for Galois rings."},{"cited_title":"MacWilliams theory over Zk an d nu-function over lattices,","cited_arxiv_id":null,"evidence_quote":"Authors' earlier complete weight enumerator MacWilliams identity for $\\mathbb{Z}_k$ that the theta-function Theorem 2 reformulates."}],"review_version":1}