{"id":"904f2861-d50e-4fdd-b776-7f0a3d2ac99f","arxiv_id":"2507.03709","paper_version":3,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The authors enumerate all semirings of order at most 6 (and some classes up to order 8), giving exact counts up to isomorphism and up to isomorphism or anti-isomorphism.","lead":"This note counts all finite semirings of up to six elements, reporting exact numbers up to isomorphism and up to isomorphism or anti-isomorphism. It also counts special subclasses like additively idempotent or commutative semirings for slightly larger sizes, giving researchers a complete census of small examples to test conjectures.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The manuscript never states how anti-isomorphic semigroup pairs are handled, so the up-to-isomorphism counts may be missing the opposite semigroup for every non-self-dual class.","rationale":"The double-coset framework is mathematically sound, and the cross-checks for n up to 5, including the careful explanation of the 16-table discrepancy with alg, give genuine support for the pipeline. The central risk lies entirely in the construction of the candidate set of semigroups. Smallsemi is described as containing semigroups 'up to isomorphism and anti-isomorphism'; for the up-to-isomorphism tables, every isomorphism class of multiplication tables is needed. If the library stores only one representative of each anti-isomorphism pair, then the opposite semigroup of every non-self-dual semigroup is missing, and since the opposite of a semiring is again a semiring, the published up-to-isomorphism counts would be systematically too small. The manuscript does not explain how this is handled, so the method as described is incomplete for the central claim, regardless of Smallsemi's correctness. The published code likely handles this, but acceptance should be contingent on confirming that both orientations are included or on an explicit statement in the paper. This is the same family of concern the reader identified, sharpened to the specific algorithmic step where Smallsemi's convention interacts with the claimed counts.","tokens_in":8650,"tokens_out":23116,"duration_ms":280705,"concrete_test":"Inspect the published Semirings source (github.com/pramothragavan/semirings) at the point where it loads semigroups from Smallsemi. Confirm that for each non-self-dual semigroup S in the library, the iteration set contains S^op, or that the code calls a Smallsemi function that returns semigroups up to isomorphism. A direct search for the dual/opposite constructor settles it. If both orientations are included, no correction is needed; if only one orientation appears, rerun the n=5 'up to isomorphism' count with the duals added and compare with the published 57,427.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central counts in Tables 1-3 'up to isomorphism' require the full list of multiplicative semigroups up to isomorphism, not merely up to anti-isomorphism. Section 2 says only that the semigroups of order at most 8 are available in Smallsemi 'up to isomorphism and anti-isomorphism', and that the algorithm iterates over this library. If Smallsemi stores a single representative of each anti-isomorphism pair (its usual convention), then for every semigroup S not anti-isomorphic to itself, the opposite semigroup S^op is absent from the iteration. Because the opposite of a semiring is again a semiring, each such omission removes at least one isomorphism class from the 'up to isomorphism' tables; the undercount would be roughly the difference between the 'up to isomorphism' and 'up to isomorphism or anti-isomorphism' columns. The paper does not state that both orientations are generated or that an 'up to isomorphism' list is used instead. Thus the method as written is incomplete for the central claims, independently of whether Smallsemi itself is complete. The same candidate-list issue is the sharpest way the dependence on Smallsemi can bite.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents a computational enumeration of finite semirings of small order. The authors prove a double-coset criterion (Theorem 1.3) that reduces the enumeration of semirings with fixed additive and multiplicative semigroups to testing distributivity for one representative of each double coset Aut(S,×)\\Sym(S)/Aut(S,+) (and the analogous criterion with Aut* for anti-isomorphism). Using the GAP packages Smallsemi and Semigroups, they produce tables of counts for n≤6 (and some classes up to n=8), including the verification that there are 866 ai-semirings of order 4, and they report a discrepancy with the alg package for order 5, which they attribute to 16 invalid multiplication tables in alg.","tokens_in":8882,"tokens_out":12210,"duration_ms":140612,"significance":"If the counts are correct, they are the first exact counts for semirings of order 6 and several classes of order 7 and 8, and the method is an elegant, parameter-free reduction that will be useful for further enumeration. The paper ships reproducible code in the GAP package Semirings, and the double-coset criterion is proved in the paper. The external agreement for orders up to 5 (with MathStructures and Ren et al.) provides strong evidence of correctness. The main risk is the reliance on the Smallsemi library for the semigroup representative lists, and the lack of an independent check for n=6 and n=7.","major_comments":[{"comment":"The text says that 'it suffices to consider every commutative semigroup (S,+) and semigroup (S,×)' and then notes that Smallsemi contains semigroups 'up to isomorphism and anti-isomorphism.' For the columns labeled 'up to isomorphism' in Tables 1-3, the multiplicative semigroups must be enumerated up to isomorphism, not merely up to anti-isomorphism. If Smallsemi stores only one representative of each anti-isomorphism pair, then for every semigroup S that is not anti-isomorphic to itself, the opposite semigroup S^op is absent from the iteration, and every semiring with multiplicative part S^op is omitted from the up-to-isomorphism counts. The disparity between the n=2 entries (10 versus 9) shows that the implementation must in fact use both orientations, or a list up to isomorphism, but the paper does not state this. Please state explicitly which representative list is used for each column and how the opposite orientation is handled, since the n=6 and n=7 counts have no external cross-check.","section":"Section 2, algorithm description"},{"comment":"The paper assumes the completeness and correctness of the Smallsemi library for semigroups of order up to 8. The cross-checks against alg and MathStructures for n≤5 are valuable, but for n=6 and n=7 there is no independent verification, and any missing isomorphism class in Smallsemi would directly change the reported counts. I do not regard this as a flaw in the method, but the paper should explicitly acknowledge this dependence, state the exact version and options used (the reference gives Version 0.7.2), and ideally provide additional validation, such as an independent generation of semigroups of order 6 up to isomorphism from a different source or a consistency check via Burnside's lemma on the number of semirings.","section":"Section 2, last paragraph before Table 1"}],"minor_comments":[{"comment":"In the reverse direction, the displayed chain of equalities contains a stray fragment 'τ α^{-1} ∈ Aut(S,×)' in the middle of the derivation; this should be re-typeset so that the reader can follow the argument without ambiguity.","section":"Section 1, proof of Theorem 1.3"},{"comment":"The phrase 'up to isomorphism and anti-isomorphism' is ambiguous; please use 'up to isomorphism or anti-isomorphism' or explicitly spell out the equivalence relation being used for the semigroup representatives.","section":"Section 2"},{"comment":"The table uses 'with 1' without an explicit definition; please state in the caption or the introduction that this means a multiplicative identity element, matching the earlier definition of 'with 0'.","section":"Table 1 caption"},{"comment":"The paper mentions that the longest computation took approximately 2400 CPU hours; for reproducibility, consider reporting the total CPU time and the exact versions of GAP, Smallsemi, and Semigroups used, and provide a small sample script that reproduces one of the table entries.","section":"Section 2, reproducibility"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a computational note whose value depends on the reproducibility of the counts. The referee's main concern is not the mathematics but the under-specification of the semigroup list used for the 'up to isomorphism' columns. The authors should also consider archiving the Semirings package (e.g., on Zenodo) with a version number, since the paper currently points only to a GitHub URL. There is no concern about novelty or scope; the note fits math.RA."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is a good, honest enumeration paper. The genuinely new content is the n=6 row of Table 1 and the up-to-anti-isomorphism counts, which to my knowledge have not appeared before. The method is elementary but effective: use Corollaries 1.4 and 1.6 to reduce the problem to iterating over double cosets for each candidate additive and multiplicative semigroup. Theorem 1.3 is correct; the proof has a small typographic slip in the (=>) direction, but the intended argument reconstructs cleanly. The external checks are reassuring: n<=5 up-to-isomorphism counts match MathStructures and the 866 ai-semirings of order 4 from Ren et al. The code is public, and the discrepancy with alg is explained concretely with a reference to a GitHub issue.\n\nThe soft spot is the one the stress-test flagged. The paper says the semigroups from Smallsemi are available \"up to isomorphism and anti-isomorphism\" without clarifying what that means for the candidate list. For the central \"up to isomorphism\" tables, you need every isomorphism class of multiplicative semigroups, not merely one representative from each anti-isomorphism pair. If Smallsemi stores only one orientation per anti-pair (which is its usual convention), then the algorithm as written would miss the semirings whose multiplicative part is the missing opposite, and the up-to-isomorphism counts would be too low. The fact that the n<=5 counts match the literature suggests the implementation actually handles this correctly--perhaps the code adds the opposite of each non-self-dual semigroup, or Smallsemi for orders <=7 stores full isomorphism-class representatives. But the paper never says which. That is an exposition gap, not necessarily a computational error, but it is load-bearing for the paper's main claim. The authors should add one sentence describing exactly what list of semigroups is iterated over and how anti-isomorphic pairs are treated.\n\nThe other limitations are minor: the n=6 and n=7 numbers are not independently verified, so confidence is moderate; and the completeness of Smallsemi is assumed without discussion. Neither is fatal for a data paper of this sort.\n\nThis is a paper for semigroup and universal algebra people who want exact counts for testing conjectures, not for readers seeking new theory. It deserves a serious referee--with a request that the candidate-list issue be clarified. I would not cite it in my own work, but I would point colleagues to it.","headline":"Solid enumeration data; the n=6 counts are new and likely correct, but the paper needs to state explicitly how the Smallsemi anti-isomorphism convention is handled.","tokens_in":714,"tokens_out":3762,"would_cite":false,"duration_ms":214285,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["16Y60","20M10","05A15"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper establishes exact enumeration counts for finite semirings of order at most 6, using a double-coset reduction that also yields selected classes up to order 8.","keywords":["semirings","finite semirings","enumeration","isomorphism classes","additively idempotent semirings","commutative semirings","double cosets","small semigroup library"],"falsifier":"Write an independent program that, on a four-element set, enumerates all pairs of binary operations satisfying commutativity, associativity, and distributivity and then quotients by relabelling; if the number of resulting isomorphism classes is not $2{,}341$, or the additively idempotent subclass is not $866$, the tables in this paper are wrong.","tokens_in":8471,"feed_emoji":"🧮","tokens_out":12945,"duration_ms":141987,"temperature":0.7,"pith_summary":"The paper establishes exact counts of finite semirings of small order: up to isomorphism and up to isomorphism-or-anti-isomorphism for every order $n\\leq 6$, with selected special classes carried to $n=7$ or $n=8$. The headline numbers are $7{,}571{,}579$ semirings of order $6$ up to isomorphism and $4{,}102{,}358$ up to isomorphism or anti-isomorphism. The enumeration rests on a double-coset criterion that tests distributivity on one representative of each double coset rather than on every relabelling, together with an existing complete library of small semigroups. A sympathetic reader would care because exact counts for unconstrained algebraic structures at these orders were largely missing, and because the paper resolves a discrepancy over the number of order-$5$ semirings.","feed_headline":"Exact semiring counts now reach order 6","feed_subtitle":"A double-coset reduction turns semiring enumeration into a finite check; special classes run to order 8.","key_machinery":"The engine is Theorem 1.3, the double-coset criterion: for a commutative additive semigroup $(S,+)$ and semigroup $(S,\\times)$, the relabelled multiplications $\\times^\\sigma$ and $\\times^\\tau$ are semigroup-isomorphic through a bijection in $\\operatorname{Aut}(S,+)$ iff $\\sigma$ and $\\tau$ belong to the same double coset in $\\operatorname{Aut}(S,\\times)\\setminus\\operatorname{Sym}(S)/\\operatorname{Aut}(S,+)$. Corollaries 1.4 and 1.6 convert this into an enumeration rule: to count semirings up to isomorphism, it suffices to walk over all commutative semigroups $(S,+)$ and all semigroups $(S,\\times)$ from the library, and to test distributivity of $+$ against $\\times^\\sigma$ for one representative $\\sigma$ of every double coset. This replaces the naive process of testing every permutation with a much smaller representative set, and the representatives are computed from the automorphism groups of the two semigroups, which are themselves obtained by reducing automorphism-group computation to graph automorphism.","core_discovery":"For a fixed commutative additive semigroup $(S,+)$ and a fixed semigroup $(S,\\times)$, every semiring on $S$ whose additive part is $+$ and whose multiplication is a relabelling of $\\times$ corresponds to a permutation $\\sigma\\in\\operatorname{Sym}(S)$. The paper proves that two such relabelled multiplications $\\times^\\sigma$ and $\\times^\\tau$ give isomorphic semirings exactly when $\\sigma$ and $\\tau$ lie in the same double coset $\\operatorname{Aut}(S,\\times)\\setminus\\operatorname{Sym}(S)/\\operatorname{Aut}(S,+)$, with an analogous statement, using $\\operatorname{Aut}^*(S,\\times)$, governing isomorphism or anti-isomorphism. The paper's central claim is that, by checking distributivity for one representative of each double coset and drawing the candidate semigroups from a precomputed library of all semigroups of order at most $8$, the totals in Tables 1--4 are the exact numbers of finite semirings up to the stated equivalences. These include $7{,}571{,}579$ semirings of order $6$ up to isomorphism, $4{,}102{,}358$ up to isomorphism or anti-isomorphism, and the confirmed value of $866$ additively idempotent semirings of order $4$ (semirings in which $x+x=x$ for every element).","pith_inferences":["A likely next step is to compute the missing order-7 total for unconstrained semirings up to isomorphism; the paper already lists the order-7 total up to isomorphism-or-anti-isomorphism ($48{,}152{,}448{,}707$), so the missing figure is plausibly within reach at comparable computational cost.","Because the cost of the method scales with the number of automorphism-group double cosets rather than with $n!$ relabellings, the same program may extend to order $8$ for classes whose semigroups have small automorphism groups; that scaling statement is an editorial guess, not a claim in the paper.","The discrepancy the paper found in an independent library suggests that other generated collections of 'semirings' should be screened for violation of distributivity before being used in downstream work, since the offending object can be an invalid multiplication table rather than a count.","The criterion treats the two operations asymmetrically only through their automorphism groups, so the same double-coset scheme could count other two-operation algebras with one commutative and one associative operation whenever an inventory of the second operation exists."],"forward_implications":["The order-6 counts, $7{,}571{,}579$ up to isomorphism and $4{,}102{,}358$ up to isomorphism or anti-isomorphism, give future enumeration algorithms a concrete numerical benchmark to match.","The corrected order-5 total of $57{,}427$ replaces the conflicting figure of $57{,}443$ that had come from another enumeration, so any database or theorem relying on the larger number needs to be rechecked.","The verification of the order-4 additively idempotent count ($866$) settles the numerical question raised in the finite-basis study that prompted this paper.","The same double-coset reduction applies to any subclass of semirings, which is why the tables also cover commutative, with-zero, with-one, and additively idempotent variants, in some cases up to order $8$."],"supporting_citations":[{"why":"It supplies the exhaustive library of semigroups of order at most 8, classified up to isomorphism and anti-isomorphism, from which all candidate multiplicative semigroups are drawn.","marker":"[11]"},{"why":"It provides the implementation of the double-coset enumeration whose output is the source of the numbers in Tables 1–4.","marker":"[16]"},{"why":"It documents the computation of double-coset representatives that the enumeration relies on.","marker":"[22]"},{"why":"It gives the reduction of semigroup automorphism computation to graph automorphism, which is how the automorphism groups of the two semigroups are obtained.","marker":"[31]"},{"why":"It supplies the automorphism-group routines applied to the semigroups during the enumeration.","marker":"[32]"},{"why":"It provides the computational system in which the enumeration and the sanity checks are run.","marker":"[19]"},{"why":"It contains the claim that there are 866 additively idempotent semirings of order 4, the verification of which initiated the paper.","marker":"[46]"},{"why":"It gives the independent enumeration whose order-5 total does not match, prompting the discrepancy analysis.","marker":"[2]"},{"why":"It records the issue in which 16 multiplication tables from the independent enumeration fail the distributivity condition, resolving the discrepancy in the paper's favour.","marker":"[3]"}],"fun_headline_variants":["Semiring census complete through order 6","Double-coset reduction yields exact semiring counts to order 6","All finite semirings of order 6 counted exactly","Exact semiring enumeration now reaches order 6"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The completeness and correctness of the precomputed library of semigroups of order at most 8 is assumed, and a missing or mislabelled isomorphism class there would make every reported semiring count too small.","fun_headline_variants_meta":{"raw":{"variants":["Semiring census complete through order 6","Double-coset reduction yields exact semiring counts to order 6","All finite semirings of order 6 counted exactly","Exact semiring enumeration now reaches order 6"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000224,"raw_usage":{"total_tokens":1415,"prompt_tokens":852,"completion_tokens":563,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":468,"completion_tokens_details":{"reasoning_tokens":498}},"tokens_in":468,"tokens_out":563,"duration_ms":6263,"temperature":1.0,"reasoning_tokens":498,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T20:03:49.692245+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Write an independent program that, on a four-element set, enumerates all pairs of binary operations satisfying commutativity, associativity, and distributivity and then quotients by relabelling; if the number of resulting isomorphism classes is not $2{,}341$, or the additively idempotent subclass is not $866$, the tables in this paper are wrong.","supporting_citations":[{"cited_title":"Distler and J","cited_arxiv_id":null,"evidence_quote":"It supplies the exhaustive library of semigroups of order at most 8, classified up to isomorphism and anti-isomorphism, from which all candidate multiplicative semigroups are drawn."},{"cited_title":"Edwards, J","cited_arxiv_id":null,"evidence_quote":"It provides the implementation of the double-coset enumeration whose output is the source of the numbers in Tables 1–4."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It documents the computation of double-coset representatives that the enumeration relies on."},{"cited_title":"Graph isomorphism, general remarks","cited_arxiv_id":null,"evidence_quote":"It gives the reduction of semigroup automorphism computation to graph automorphism, which is how the automorphism groups of the two semigroups are obtained."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It supplies the automorphism-group routines applied to the semigroups during the enumeration."},{"cited_title":"The finite basis problem for additively idempotent semirings of order four, I","cited_arxiv_id":null,"evidence_quote":"It contains the claim that there are 866 additively idempotent semirings of order 4, the verification of which initiated the paper."},{"cited_title":"Bauer.alg: Algebraic library for constructive mathematics.https://github.com/andrejbauer/alg","cited_arxiv_id":null,"evidence_quote":"It gives the independent enumeration whose order-5 total does not match, prompting the discrepancy analysis."},{"cited_title":"Bauer.Issue #16: Incorrect results.https://github.com/andrejbauer/alg/issues/16","cited_arxiv_id":null,"evidence_quote":"It records the issue in which 16 multiplication tables from the independent enumeration fail the distributivity condition, resolving the discrepancy in the paper's favour."}],"review_version":1}