{"id":"71e8f591-e3de-4658-adc9-b511e26372b2","arxiv_id":"1908.03228","paper_version":3,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"For primes p>q, all skew braces of order pq are explicitly constructed: only the trivial brace if p is not 1 modulo q, and exactly 2q+2 braces if p is 1 modulo q.","lead":"This paper constructs and lists every skew brace of size pq, where p and q are distinct primes. It settles a classification problem relevant to the Yang-Baxter equation and shows the count matches an independent enumeration.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified; completeness rests on Byott's external regular-subgroup classification, but no internal gap found.","rationale":"This is a clean classification argument. I re-examined the key orbit arguments: for cyclic additive group all subgroups G_b are conjugate by powers of ψ; for the non-abelian additive group the orbit representatives G_{a,0} and G_{0,d} are distinguished by the invariant parameters a and d in Propositions 3.8 and 3.11; and the orbit counts p, p(q-1), p(q-2)+1 match the e' numbers quoted from Byott. I also spot-checked the multiplication formulas for small primes, including q=2 and q=3, for associativity and left distributivity, and no inconsistency surfaced. The only place where a missing isomorphism class could enter is the completeness of Byott's four regular-subgroup families, exactly the assumption the reader identified as weakest. Since this is a published, independently usable classification and the final counts agree with the independent Byott-Alabdali enumeration, the external dependence is not a ground for changing the verdict. No internal error, no unproved necessary step beyond a routine citation, and no evidence of a missing brace emerged.","tokens_in":9292,"tokens_out":22847,"duration_ms":240867,"concrete_test":"Independently enumerate all skew braces of order pq for several small pairs with p ≡ 1 mod q, e.g. (p,q) = (3,2), (7,3), (11,5), (13,3), (31,5), using the Guarnieri-Vendramin algorithm in GAP: build the additive groups of order pq, search regular subgroups of their holomorphs, and compute isomorphism classes. Compare the isomorphism-class count with 2q+2 and identify each brace with one of the four theorem families. If the count and family identification match for all tested pairs, the inherited completeness of Byott's classification is confirmed empirically; a mismatch would pinpoint a missing or extra family.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that every skew brace of order pq appears exactly once inherits its completeness from Byott's classification of regular subgroups of Hol(C) and Hol(M), quoted as Lemmas 3.2, 3.5, 3.7 and 3.10. If a regular subgroup were missing from one of those four families, the corresponding skew brace would be missing from the theorem, so this is the single most load-bearing assumption. I found no internal reason to doubt it: the orbit computations in Propositions 3.3, 3.8 and 3.11 are internally consistent, the resulting orbit counts sum to 2q+2, and the operations in Theorems 3.4, 3.6, 3.9 and 3.12 pass associativity and distributivity checks on small examples. The 'straightforward to verify' checks are backed by the regular-subgroup construction and are not load-bearing. The paper also records agreement with Byott-Alabdali's independent enumeration for squarefree order. Thus the only genuine vulnerability is an external, published classification rather than a defect in this paper's argument.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper classifies all skew left braces of order pq, where p>q are primes, by combining Byott's classification of regular subgroups of holomorphs with the Guarnieri–Vendramin correspondence between skew braces and regular subgroups. The main theorem states that if p is not congruent to 1 modulo q, then the only skew brace of size pq is the trivial one; if p is congruent to 1 modulo q, then there are exactly 2q+2 skew braces up to isomorphism. These are described explicitly: two braces with abelian additive group Z_p × Z_q (one trivial, one bi-skew) and 2q braces with non-abelian additive group Z_p ⋊_g Z_q (one trivial, one with |ker λ|=q, q−1 bi-skew braces A_γ, and q−1 braces A_μ with ker λ=0). The proof proceeds by listing regular subgroups from Byott's work, computing their orbits under the action of the automorphism group, and then translating the orbit representatives into explicit brace operations using the holomorph correspondence.","tokens_in":9431,"tokens_out":23676,"duration_ms":223083,"significance":"If correct, this solves a natural and explicitly posed problem (Vendramin's Problem 2.15 in [24]). The result is clean and the explicit formulas for the operations are useful for further study of skew braces of small order. The paper also demonstrates a successful transfer of Hopf–Galois classification results into the skew-brace setting, and it includes a consistency check with the independent enumeration of Byott and Alabdali for squarefree order. The orbit computations are explicit and internally consistent, and the total orbit count matches the claimed 2q+2. The main external dependency is Byott's classification of regular subgroups, which is a published result; the paper's own contribution is the translation into brace language and the orbit analysis, which appears sound.","major_comments":[],"minor_comments":[{"comment":"The parametrization by γ and μ is ambiguous: the proofs define γ = (a+1)/a and μ = (d+1)/d as rational numbers, while the statements list '1<γ≤q' and '1<μ≤q'. Since g has multiplicative order q modulo p, the exponent should be understood modulo q, and γ and μ should be identified with the unique integer representatives in {2,...,q}, with q representing the residue 0. I recommend adding a sentence to clarify this point.","section":"Main Theorem; Theorems 3.9 and 3.12"},{"comment":"In the proofs of Theorems 3.6 and 3.12, the brace axioms are relegated to 'straightforward to verify' or 'it is easy to check'. Since the operations are derived from regular subgroups via Theorem 1.3, the brace property follows automatically from the regular subgroup construction; the authors could state this explicitly to avoid the impression of a gap.","section":"Theorems 3.6 and 3.12"},{"comment":"The phrase 'as j g^{d+1}-1/(g-1) runs from 0 to p' is imprecise. The intended meaning is that as j runs through the residues modulo p, the coefficient runs through all p residues when d ≠ q−1, and is identically 0 when d = q−1; this should be phrased more carefully.","section":"Proposition 3.11"},{"comment":"The term 'Burnside number' is used without definition. I suggest adding a parenthetical definition, e.g., gcd(n, φ(n)) = 1, to make the argument self-contained.","section":"Proposition 3.1"},{"comment":"Some cited works are preprints (e.g., [1] and [10]); the authors should update these references if they have appeared in final form.","section":"References"}],"recommendation":"minor_revision","confidential_remarks":"I found no load-bearing mathematical errors. The completeness of the classification is conditional on Byott's external classification of regular subgroups [7], which is a published result; the paper's own orbit computations are consistent and the orbit counts sum correctly to 2q+2. The main presentational issue is the ambiguous parametrization by γ and μ, which should be clarified, along with a few other small exposition points. The agreement with Byott–Alabdali's independent enumeration further strengthens confidence. I recommend minor revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The short version: this paper answers Vendramin's problem for skew braces of size pq, and it does so in a workmanlike way. The genuinely new contribution is the classification of the non-abelian additive type braces: when p ≡ 1 mod q, there are 2q non-abelian braces, split into q-1 with kernel size p and q-1 with trivial kernel. Add the trivial braces and one cyclic-type brace, and you get the 2q+2 total. The explicit formulas in Theorems 3.9 and 3.12 are new, and the count matches the independent enumeration by Byott and Alabdali.\n\nWhat the paper does well: it uses the holomorph correspondence seriously. Instead of guessing operations, the authors take Byott's regular subgroups of Hol(M) and compute orbits under conjugation by Aut(M). The orbit computations in Propositions 3.8 and 3.11 are explicit and easy to follow. The paper is clear about which parts come from Rump and which parts are new. The cyclic-type part is a reproof of known results, and the p not congruent to 1 case is a one-liner from Burnside's theorem, but both are needed for a complete statement.\n\nSoft spots, in order of importance. First, the completeness of the classification is imported from Byott's 2004 paper on Hopf-Galois structures (Lemmas 3.2, 3.5, 3.7, 3.10). The paper does not reprove those classifications. That is not a defect if you trust Byott, and the authors are transparent, but anyone using the result should know that the entire list depends on that external source. Second, Theorems 3.6 and 3.12 contain 'straightforward to verify' checks of the brace axioms. In both cases the operation comes from a regular subgroup, so the axioms follow from the general correspondence, but the reader has to chase that through. I'd call this minor. I found no circularity, and the agreement with Byott-Alabdali is an independent cross-check.\n\nWho is this for: people working on classification of skew braces and Hopf-Galois structures. It is a clean application of a known method to a small order family. It deserves a serious referee. No fatal problems that I can see. Have the referee focus on the completeness reduction and the orbit computations in the non-abelian case. I expect it to pass.\n\nRecommendation: send to peer review.","headline":"Clean, honest classification of skew braces of order pq; the new non-abelian case is done via explicit orbit computations, and the completeness rests transparently on Byott's published classification.","tokens_in":10006,"tokens_out":3825,"would_cite":true,"duration_ms":37500,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["16T25","20D45","16T05"],"pacs":[],"model":"deepseek-v4-flash","headline":"For primes p>q, this paper classifies all skew braces whose additive group has order pq: only the trivial brace unless p≡1 mod q, in which case exactly 2q+2 braces exist.","keywords":["skew brace","Yang-Baxter equation","Hopf-Galois extension","regular subgroup","holomorph","semidirect product","group of order pq","bi-skew brace"],"falsifier":"Fix primes p>q with p\\equiv 1\\pmod q and directly enumerate all regular subgroups of Hol(Z_p\\times Z_q) and Hol(Z_p\\rtimes_g Z_q). The claim is false if any such subgroup is not conjugate to one of the families displayed in Lemmas 3.2, 3.5, 3.7, or 3.10, or if the number of Aut-conjugacy orbits differs from 2q+2.","tokens_in":9043,"feed_emoji":"🧮","tokens_out":5686,"duration_ms":56432,"temperature":0.7,"pith_summary":"This paper completes the classification of all skew left braces whose additive group has order pq, where p>q are primes. It proves that when p is not congruent to 1 modulo q, the only skew brace of size pq is the trivial one. When p is congruent to 1 modulo q, it gives a complete list of 2q+2 skew braces up to isomorphism, with explicit operations for every brace in the list. The result matters because each skew brace yields set-theoretical solutions to the Yang–Baxter equation, and this settles one of the open classification problems in that area.","feed_headline":"All skew braces of order pq: 2q+2, or just one","feed_subtitle":"If p≡1 mod q there are exactly 2q+2 skew braces; if not, only the trivial one.","key_machinery":"The central object is the holomorph Hol(A)=A\\rtimes\\mathrm{Aut}(A) of a group A, along with the Guarnieri–Vendramin correspondence: skew braces with additive group A are in bijection with regular subgroups of Hol(A), and isomorphism classes correspond to orbits of these subgroups under conjugation by \\mathrm{Aut}(A). The paper imports Byott's explicit classification of the regular subgroups of Hol(C) and Hol(M) for groups of order pq, then computes the Aut(A)-conjugacy orbit representatives. The operation formulas are obtained from the map \\lambda_a(b)=-a+a\\circ b, which records the image of the regular subgroup under the projection to \\mathrm{Aut}(A). A separate construction shows that certain semidirect products with commuting automorphism images yield bi-skew braces.","core_discovery":"The main theorem states that for primes p>q, if p\\not\\equiv 1\\pmod q there is exactly one skew brace of size pq, the trivial one. If p\\equiv 1\\pmod q, there are exactly 2q+2 skew braces up to isomorphism, and the paper writes down their operations explicitly. The additive group is either the abelian group Z_p\\times Z_q or the non-abelian semidirect product Z_p\\rtimes_g Z_q, where g is a fixed element of order q modulo p. On the abelian side there are two braces (the trivial one and one non-trivial bi-skew brace); on the non-abelian side there are the trivial brace, one brace with |\\ker(\\$\\lambda$)|=q, q-1 braces with |\\ker(\\$\\lambda$)|=p, and q-1 braces with trivial kernel, each given by explicit formulas. The paper claims every skew brace of order pq appears exactly once in this list.","pith_inferences":["The orbit computations are self-contained once Byott's regular-subgroup families are fixed, so the classification can be verified mechanically for any concrete pair p,q by enumerating conjugacy classes inside Hol(A).","Because the same regular subgroups classify Hopf–Galois structures, the orbit analysis here also yields a count of Hopf–Galois structures of degree pq, although the paper presents it only in the language of skew braces.","Specialising to q=2 gives a concrete prediction: for any odd prime p, there should be exactly 6 skew braces of order 2p; this is a quick finite check that a reader could carry out independently.","The paper leaves implicit which of the 2q+2 isomorphism classes admit a bi-skew or inverse brace structure; a natural follow-up is to label each class by these additional properties."],"forward_implications":["Every non-degenerate set-theoretic solution to the Yang–Baxter equation whose associated skew brace has size pq is realised by one of the listed braces, once the solution-to-skew-brace reduction is applied.","For p\\not\\equiv 1\\pmod q, there are no nontrivial skew brace structures of order pq: any skew brace of that size is the trivial one.","For p\\equiv 1\\pmod q, the census size depends only on q, not on p: there are always 2q+2 braces, regardless of how large p is.","Among the 2q+2 braces, two are of cyclic additive type, and the remaining 2q are of non-abelian type, with the paper giving explicit bi-skew or non-bi-skew distinctions where relevant.","The explicit formulas allow any given skew brace of order pq to be checked against the list without searching through holomorphs."],"supporting_citations":[{"why":"Supplies Byott's explicit classification of regular subgroups of Hol(C) and Hol(M) for degree pq, which the paper uses as the starting point for orbit enumeration.","marker":"[7]"},{"why":"Establishes the correspondence between skew braces over a group A and regular subgroups of Hol(A), including the orbit parametrisation of isomorphism classes.","marker":"[15]"},{"why":"Provides the Burnside-number argument quoted for the uniqueness of the trivial skew brace when p\\not\\equiv 1\\pmod q.","marker":"[22]"}],"fun_headline_variants":["Skew braces of size pq: trivial only, or 2q+2","For primes p>q, count skew braces: 1 or 2q+2","All skew braces of order pq are now known","Explicit skew braces for order pq: 2q+2 if p≡1 mod q","Skew braces of pq: a full classification"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The completeness of the final list rests entirely on Byott's classification of the regular subgroups of Hol(C) and Hol(M) for groups of order pq, quoted in Lemmas 3.2, 3.5, 3.7, and 3.10; if that classification missed any regular subgroup, the corresponding skew brace would be missing from the theorem.","fun_headline_variants_meta":{"raw":{"variants":["Skew braces of size pq: trivial only, or 2q+2","For primes p>q, count skew braces: 1 or 2q+2","All skew braces of order pq are now known","Explicit skew braces for order pq: 2q+2 if p≡1 mod q","Skew braces of pq: a full classification"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000265,"raw_usage":{"total_tokens":1553,"prompt_tokens":837,"completion_tokens":716,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":453,"completion_tokens_details":{"reasoning_tokens":619}},"tokens_in":453,"tokens_out":716,"duration_ms":7115,"temperature":1.0,"reasoning_tokens":619,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:21:55.380517+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Fix primes p>q with p\\equiv 1\\pmod q and directly enumerate all regular subgroups of Hol(Z_p\\times Z_q) and Hol(Z_p\\rtimes_g Z_q). The claim is false if any such subgroup is not conjugate to one of the families displayed in Lemmas 3.2, 3.5, 3.7, or 3.10, or if the number of Aut-conjugacy orbits differs from 2q+2.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies Byott's explicit classification of regular subgroups of Hol(C) and Hol(M) for degree pq, which the paper uses as the starting point for orbit enumeration."},{"cited_title":"Smoktunowicz and L","cited_arxiv_id":null,"evidence_quote":"Provides the Burnside-number argument quoted for the uniqueness of the trivial skew brace when p\\not\\equiv 1\\pmod q."}],"review_version":1}