{"id":"06267f9c-d99e-4b43-a8f7-dcc9e4a503a9","arxiv_id":"2501.19371","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"At most 13 real quadratic fields admit a ternary universal quadratic lattice, and for several of these fields explicit universal lattices are constructed.","lead":"This paper proves that only 13 real quadratic fields can possibly have a universal ternary quadratic lattice, and it gives explicit discriminant bounds that make the determination possible. It is a major step toward Kitaoka's Conjecture, and for 5 of the 13 fields it constructs genuinely universal lattices.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 13-field list in Theorem 1.1 depends on unshipped Magma exclusions in Propositions 7.4 and 7.5; without code, data, or certificates, the decisive finite screening is not independently verifiable.","rationale":"I read the paper in good faith and traced the structure of the main argument. The analytic core, Theorem 1.2 and its proof via Theorems 3.1, 5.1, 5.4, 6.2, 6.3, and 6.6, is elaborate but appears internally consistent: the inequalities in Corollary 3.3 and the rank arguments in Sections 5 and 6 use standard lattice and determinant facts, and I did not find a clear mathematical contradiction. The reduction from a universal ternary lattice to a classical lattice representing 2O_F^+ is valid, and the resulting bound ∆D < 10000 is correctly specialized to D < 10000 for D ≡ 1 (mod 4) and D < 2500 for D ≡ 2, 3 (mod 4). The load-bearing step is therefore the finite computational exclusion of all other discriminants. The reader's weakest_assumption identifies exactly this point. The computations are finite and checkable, but the manuscript does not ship the Magma code, state the search algorithm in detail, or provide certificates. In particular, Propositions 7.4 and 7.5 split the discriminants into several branches with different test sets, and the proof relies on the exhaustive correctness of the 'simple program in Magma'. This is precisely the kind of missing support that should be flagged. It does not warrant rejection, because the claimed computations are plausible and the surrounding theory is strong; but it does warrant a conditional verdict pending release of code and independent verification. I therefore recommend no change to the reader's CONDITIONAL verdict.","tokens_in":34904,"tokens_out":6971,"duration_ms":73842,"concrete_test":"Request the Magma scripts for Propositions 7.4 and 7.5 and independently re-run the exclusion using a separate exact-arithmetic enumeration (e.g., Sage or PARI) for every relevant D up to the stated bounds. For each squarefree D < 10000 with D ≡ 1 (mod 4) outside the eight listed values, and each squarefree D < 2500 with D ≢ 1 (mod 4) outside {2,3,6,7,10}, check whether a totally positive semidefinite rank-at-most-3 matrix with diagonal S_D (or the relevant four- or eight-element test set) exists in M4(1/2O_F), and compare the resulting exceptional lists with the 22-value and 26-value lists in Propositions 7.4 and 7.5. If the independent implementation reproduces every exclusion, the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"To pass from the analytic bound of Theorem 1.2 to the complete list in Theorem 1.1, the proof must exclude every squarefree D < 10000 with D ≡ 1 (mod 4) outside the eight listed fields, and every D < 2500 with D ≢ 1 (mod 4) outside {2,3,6,7,10}. This exclusion is performed entirely by the Magma computations summarized in Propositions 7.4 and 7.5: the text says 'one can check' and 'we checked this by running a simple program in Magma', while the code is only 'available upon request' (Section 7). The search is in principle finite, since a positive semidefinite matrix with fixed diagonal has off-diagonal entries bounded by the Cauchy–Schwarz inequality in each embedding, so an exact enumeration over M4(1/2O_F) is possible. However, the preprint gives no algorithm, source code, output data, or certificate of correctness. A bug in the enumeration—for example, an incorrect bound on off-diagonal entries, an incomplete treatment of non-free lattices, or an error in the exceptional-field branching—would change the final list. The theoretical material in Sections 3–6 is detailed and appears internally consistent, but the central claim's conclusive step is a computational black box that the manuscript does not make reproducible.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves that there are at most thirteen real quadratic fields admitting a ternary universal quadratic lattice, namely D = 2, 3, 5, 6, 7, 10, 13, 17, 21, 33, 41, 65, 77 (Theorem 1.1 / Theorem 7.1). The proof combines a general analytic and local argument bounding the discriminant of a real quadratic field that admits a classical lattice representing all totally positive multiples of a fixed integer m (Theorems 1.2 and 3.1, with ranks 3 through 7 treated in Sections 4–6) with a finite computer search over all remaining small discriminants (Propositions 7.4 and 7.5). The paper also gives a conjectural complete list of universal ternary lattices over the ten non-classical fields (Theorem 7.6), proves universality for several of them via class number 1 and local-global arguments (Theorem 7.8), and discusses separations among minimal ranks of universal forms with various restrictions (Remarks 7.9 and 7.10).","tokens_in":1,"tokens_out":3396,"duration_ms":130012,"significance":"If the finite computer checks are correct, this is a complete and explicit solution of Kitaoka's Conjecture for real quadratic fields, a substantial advance over the earlier finiteness result of Kim–Kim–Park. The theoretical discriminant bounds for lattices representing m-times the totally positive elements are new and carefully derived, use explicit constants from Burgess, Treviño, and GRH results, and are not fitted to the final list. The paper also gives concrete candidate lattices, proves universality of several of them by class-number-one local-global arguments, and resolves several informal conjectures about differences between classical, non-classical, diagonal, and free universal forms. The main weakness is that the decisive exclusion of about ten thousand discriminants rests on Magma computations that are described only verbally, with no code, input data, or output data included; this part is not independently verifiable as submitted.","major_comments":[{"comment":"The complete list in Theorem 1.1 depends crucially on the finite exclusions in Propositions 7.4 and 7.5. The proofs say only that 'one can check' and 'we checked this by running a simple program in Magma', while the code is 'available upon request' (page 21). No algorithm, source code, output data, or certificates are provided. Since a bug in the enumeration of positive semidefinite matrices in M4(1/2O_F)—for example in the bound on off-diagonal entries, the treatment of non-free lattices, or the special-case branching for the listed exceptional discriminants—would change the final list, the central claim is not reproducible in its current form. Please supply the Magma code, a precise description of the finite enumeration and its bounds, and the output data (or verifiable certificates) for all excluded D, preferably as ancillary files.","section":"Section 7, Propositions 7.4 and 7.5"},{"comment":"The classification of all universal ternary lattices over the ten fields D = 6, 7, 10, 13, 17, 21, 33, 41, 65, 77 and the assertion that each listed lattice represents all totally positive elements of norm at most 250000 are again justified only by 'we checked' statements with code available upon request. In particular, the claim that no free universal ternary lattice exists for D = 10 and D = 65 relies on the unverified computation behind Theorem 7.6. These computations should be made available in the same reproducible form as those for Propositions 7.4 and 7.5.","section":"Section 7, Theorem 7.6 and Theorem 7.8"},{"comment":"I found no internal inconsistency in the theoretical rank bounds: the reduction from Theorem 3.1 to Corollary 3.3 and Theorem 1.2 is coherent, and the estimates for C1, D1, C2, D2 using Lemma 3.4 and the Bennett table are consistent with the stated exponents. The main theoretical argument appears sound, but it establishes only a conditional reduction to a finite check; the completeness of the final list is therefore entirely contingent on the missing computational evidence described above.","section":"Sections 3–6"}],"minor_comments":[{"comment":"The author line contains an apparent rendering artifact: 'B/suppress LA˙ZEJ ˙ZMIJA' should be corrected to 'Błażej Żmija'.","section":"Title page / author list"},{"comment":"The table headers 'for m = 1 for m = 2' are ambiguous; please separate the columns clearly and state in the caption that all entries are upper bounds for the discriminant ΔD.","section":"Theorem 1.2, table"},{"comment":"In the displayed line for the m ≥ 3 unconditional case of rank 6, the expression '≪ m^{55/8}(log m)^{13/4}' is correct up to the notation used, but the intermediate inequality involving max{...} would be easier to follow if the discarded first term were explicitly justified by a comparison of the exponents.","section":"Section 3, proof of Theorem 1.2"},{"comment":"Remark 4.7 states that γ2 can be taken to be 5 in several places, while the paper globally sets γ2 = 7; this is harmless but could confuse readers. Please clarify that the global convention is retained for uniformity and indicate precisely where the sharper value is used.","section":"Remark 4.7"},{"comment":"References [Ki3] and Krasenský–Scharlau are cited as unpublished/in preparation; this is acceptable, but they should be marked as such in the bibliography, and the dependence on them should be stated clearly.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The theoretical parts of the paper are solid and the result is important, but the completeness of Theorem 1.1 currently rests on an unshipped computational black box. I would be willing to accept after the authors provide the Magma code, a precise statement of the verified enumeration, and the output data for Propositions 7.4 and 7.5 (and ideally for Theorem 7.6/7.8). If the code or data reveal an error, the main list would change, so this is a load-bearing issue rather than a presentation matter."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper is exactly what the reader says: a serious, mostly theoretical advance with one reproducible-computations gap. The abstract's claim is backed by a detailed argument; the 13-field list is new and, if the computations are right, settles Kitaoka's Conjecture for quadratic fields in effective form.\n\nThe genuinely new and impressive parts are the discriminant bounds in Theorem 1.2 and the rank-separation results in Remark 7.9. The proof idea – represent m, 2m, and then a third multiple forced by the least quadratic non-residue – is executed with care. I read Sections 4–6 for internal consistency and found no contradiction. There are no free parameters fitted to the final list; the bounds come from Burgess/Treviño, and the local arguments are clean.\n\nThe soft spot is exactly where the reader puts it. Propositions 7.4 and 7.5 exclude roughly 10^4 fields, and the text says 'one can check' and 'we checked this by running a simple program in Magma', with code 'available upon request'. No code, no output data, no certificates. A bug in the enumeration, say in the off-diagonal bound or the handling of non-free lattices, would change the main theorem. The search is finite and the bounds are explicit, so the gap is repairable, but it is a genuine reproducibility gap in the central claim. The same applies to the candidate lists in Theorem 7.6 and the norm-limited universality checks in Theorem 7.8.\n\nI want to be clear: this is not a case of a paper where the computations are an afterthought. They are load-bearing. The authors are honest that Conjecture 7.2 is conjectural, and the proof of Theorem 1.1 doesn't use it. That helps. But a referee cannot verify the main list without rerunning the Magma code or inspecting it.\n\nBottom line: deserves a serious referee. I would send it out with a request that the code and output be posted (e.g., on arXiv or a repository) before acceptance. If the computations check out, this is a significant result and I'd likely cite it in my own work. Reading group? Maybe – the theoretical sections are worth discussing, but the computational gap will dominate the conversation.","headline":"Effective resolution of Kitaoka's conjecture for quadratic fields with a reproducible-computations caveat: theoretical core is strong, but the 13-field list depends on unshipped Magma output.","tokens_in":35725,"tokens_out":3062,"would_cite":true,"duration_ms":30038,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["11E12","11E20","11E25","11R04","11R11","11R80"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that at most 13 real quadratic fields admit a universal ternary quadratic lattice.","keywords":["universal quadratic form","quadratic lattice","ternary lattice","real quadratic field","Kitaoka's conjecture","least quadratic non-residue","discriminant bound"],"falsifier":"Re-run the exhaustive computer search for any $D$ outside the thirteen values, e.g. $D=14$, checking whether any ternary lattice represents the set $\\{1,2,\\alpha_1,\\alpha_2\\}$ defined in Proposition 7.4; the discovery of such a lattice would disprove Theorem 1.1. Independently, finding a totally positive element of $Q(\\sqrt{13})$ not represented by the candidate form $\\phi_{24}^{(13)}$ would refute the claimed universality of Theorem 7.8.","tokens_in":34768,"feed_emoji":"🧮","tokens_out":11932,"duration_ms":106277,"temperature":0.7,"pith_summary":"Kitaoka's conjecture predicts that only finitely many totally real number fields can have a universal ternary quadratic lattice, one that represents every totally positive algebraic integer. This paper proves the conjecture for real quadratic fields in explicit form: if $Q(\\sqrt{D})$ carries a universal ternary lattice, then $D$ must be one of thirteen numbers, namely $2, 3, 5, 6, 7, 10, 13, 17, 21, 33, 41, 65,$ and $77$. For each of these thirteen fields the paper produces a candidate universal ternary lattice, and for five of them—$D = 2, 3, 5, 13, 17$—universality is actually established. As a by-product, the paper gives explicit upper bounds on the discriminant of any real quadratic field whose classical lattice of rank at most $7$ represents all totally positive multiples of a fixed integer $m$. The result turns a qualitative finiteness statement into a finite, checkable list.","feed_headline":"Only 13 real quadratic fields can host a universal ternary lattice","feed_subtitle":"The explicit list of thirteen discriminants comes with candidate lattices, settling Kitaoka's conjecture for quadratic fields.","key_machinery":"The engine of the proof is a rank-growth machine built from vectors of prescribed values. Starting from linearly independent vectors $v_1, v_2$ representing $m$ and $2m$, Lemma 4.4 shows that one can choose a small prime $p \\le 2m^2$ and an integer $2 \\le C_1 \\le \\gamma_p$ (the least quadratic non-residue modulo $p$) such that no vector representing $C_1 m$ can lie in the subspace spanned by $v_1, v_2$; a vector $v_3$ representing $C_1 m$ is therefore automatically independent. Iterating this 'independence by multiplier' argument with further multipliers $C_2, D_1, D_2$ produces vectors $v_0, v_1, \\dots, v_7$ whose Gram matrix is forced to be positive definite, which in turn forces the lattice rank to exceed any given bound. The discriminant bounds of Theorem 1.2 are obtained by combining these rank-forcing constraints with Hadamard's inequality on the determinants of Gram matrices, while the final exclusion of all but thirteen discriminants is a finite computer search over the finitely many ternary lattices that could represent a small explicit set of totally positive elements.","core_discovery":"On its own terms, the paper's central discovery is Theorem 1.1: if a real quadratic field $Q(\\sqrt{D})$ admits a universal ternary quadratic lattice, then $D$ belongs to the finite set $\\{2,3,5,6,7,10,13,17,21,33,41,65,77\\}$. This is a strong version of Kitaoka's conjecture because it gives the complete list of possible discriminants rather than merely a finiteness assertion. The proof rests on Theorem 1.2, which bounds the discriminant of any classical totally positive definite lattice of rank at most $7$ over $Q(\\sqrt{D})$ that represents every element of $mO_F^+$ by explicit functions of $m$ and the rank; for $m=2$ and rank $3$ this gives $\\Delta_D < 10000$. The list is then completed by a finite computation that excludes all remaining discriminants, and by the exhibit of candidate lattices for the thirteen survivors. The authors further prove that $Q(\\sqrt{10})$ and $Q(\\sqrt{65})$ admit no free universal ternary lattice, so only eleven of the thirteen fields can support a universal ternary quadratic form, and they list the conjecturally complete set of universal ternary lattices for the ten fields with $D>5$, proving universality for the fourteen lattices with class number one.","pith_inferences":["The rank-growth lemmas use only the quadratic nature of the field, so analogous explicit discriminant bounds should hold for totally real fields of fixed higher degree, with the unit group arithmetic entering only through the final finite check.","The list of thirteen discriminants is likely the complete truth: if Conjecture 7.2 is proven, then every real quadratic field admitting a universal ternary lattice has D among these values, and the only remaining work is to verify universality of the listed candidate lattices.","The non-free lattices over $Q(\\sqrt{10})$ and $Q(\\sqrt{65})$ suggest that a nontrivial class group does not obstruct universal ternary lattices, although it does obstruct free ones; testing other small discriminants with nontrivial class groups could reveal further examples."],"forward_implications":["Kitaoka's conjecture is resolved for real quadratic fields with an explicit list of thirteen possible discriminants, replacing a qualitative finiteness result with a finite set that can be checked directly.","Universal ternary lattices provably exist over $Q(\\sqrt{2}), Q(\\sqrt{3}), Q(\\sqrt{5}), Q(\\sqrt{13})$, and $Q(\\sqrt{17})$, so the minimal rank of a universal lattice is 3 for these fields.","The minimal rank of a classical universal form over $Q(\\sqrt{13})$ is 4, and of a diagonal universal form is at least 5, showing that these invariants genuinely differ.","For $m=2$ and rank 3, the discriminant bound $\\Delta_D < 10000$ is sharp enough to make the full classification of ternary lattices representing $2O_F^+$ computationally feasible."],"supporting_citations":[{"why":"Establishes the classical base case: the only real quadratic fields with a classical universal ternary form are $D=2,3,5$, which the present paper generalizes to non-classical lattices.","marker":"[CKR]"},{"why":"Proves the finiteness of real quadratic fields admitting a universal lattice of rank at most 7, providing the qualitative starting point that the explicit bounds refine.","marker":"[KKP]"},{"why":"Supplies the explicit unconditional bound on the least quadratic non-residue $\\gamma_p$ that makes the discriminant bounds for $m=1,2$ explicit.","marker":"[Tre]"},{"why":"Gives the standard theory of quadratic lattices and p-adic square classes on which the rank-growth lemmas and local considerations rest.","marker":"[OM]"},{"why":"The computer algebra system in which the finite exclusion computation for the discriminants below the explicit bound is implemented.","marker":"[BCP]"}],"fun_headline_variants":["Kitaoka's conjecture proven for quadratic fields","Exactly 13 quadratic fields have universal ternary lattices","Explicit list: 13 quadratic fields with universal ternary lattices","Thirteen discriminants settle Kitaoka's conjecture for quadratic fields","Kitaoka conjecture settled for quadratic fields"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the finite computer search—described in the paper but not shipped as code—correctly proves, for every excluded discriminant below the explicit bound, that no ternary lattice can represent the specified finite set of totally positive elements.","fun_headline_variants_meta":{"raw":{"variants":["Kitaoka's conjecture proven for quadratic fields","Exactly 13 quadratic fields have universal ternary lattices","Explicit list: 13 quadratic fields with universal ternary lattices","Thirteen discriminants settle Kitaoka's conjecture for quadratic fields","Kitaoka conjecture settled for quadratic fields"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000935,"raw_usage":{"total_tokens":3963,"prompt_tokens":868,"completion_tokens":3095,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":484,"completion_tokens_details":{"reasoning_tokens":3016}},"tokens_in":484,"tokens_out":3095,"duration_ms":25020,"temperature":1.0,"reasoning_tokens":3016,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-09T20:19:41.213097+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-run the exhaustive computer search for any $D$ outside the thirteen values, e.g. $D=14$, checking whether any ternary lattice represents the set $\\{1,2,\\alpha_1,\\alpha_2\\}$ defined in Proposition 7.4; the discovery of such a lattice would disprove Theorem 1.1. Independently, finding a totally positive element of $Q(\\sqrt{13})$ not represented by the candidate form $\\phi_{24}^{(13)}$ would refute the claimed universality of Theorem 7.8.","supporting_citations":[],"review_version":1}