{"id":"691d626f-d163-42da-a72f-1953e2d658f0","arxiv_id":"2508.19418","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A complete enumeration algorithm for finite real hyperfields with cyclic positive cones, with classification data to order 15, C-characteristics to order 17, and many new non-Krasner quotient hyperfields.","lead":"The authors present an algorithm that enumerates all finite real hyperfields with cyclic positive cones, a class of algebraic structures generalizing fields, and they classify these up to size 15 while identifying an invariant for size 17. This is a computational advance in hyperfield theory and supplies many new examples that cannot be obtained from the classical quotient construction.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"N=8 C-characteristic enumeration is unsupported: Section 6 restricts to C-char>2 without describing an exhaustive search or proof, so Table 6.2's 190 classes and 'no higher C-characteristics' are not reproducible from the paper.","rationale":"I agree with the reader's weakest-assumption diagnosis. Theorem 3.3 and Theorem 4.3 give a parameterization that makes Algorithm 14 exhaustive for a fixed N in principle, and the small-N counts are consistent with the stated case counts. The load-bearing gap is the N=8 row: the paper changes the search scope without giving the optimized procedure or a completeness proof. Because 255^5 is about 10^12, the 190-class table cannot be verified from the description in Sections 5-6. This is a correctness risk for the order-17 identification, not for the N<=7 classification. I also note that Remark 3.4's minimality assertion is unproved, but it is secondary to the classification claim and could be checked by small counterexample searches. The appropriate verdict remains CONDITIONAL, so no adjustment to the reader's verdict is needed.","tokens_in":20866,"tokens_out":14889,"duration_ms":140532,"concrete_test":"Obtain or independently reimplement the N=8 search with a published optimized exact method: for example, backtracking over the (K+1)-tuple parameterization with Even_Cond and (kr0)-(kr2) pruning, or a SAT/CP encoding of the hyperfield axioms, enumerating all C-char>2 classes. Verify that exactly 190 isomorphism classes with C-char=3 are found and that no class with C-char>3 exists. If the enumerated count differs, or if no such search is feasible without additional restrictions, Table 6.2's completeness claim should be downgraded to a partial computation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's strongest contribution for order-17 hyperfields is the Section 6 claim that restricting to C-characteristic >2 yields exactly 190 isomorphism classes with C-char=3 and no higher C-characteristics. The only exhaustive algorithm supplied is brute-force Algorithm 14, which for N=8 would enumerate (2^8-1)^5 ~ 1.08e12 candidate 5-tuples before filtering. Section 6 states the restriction but provides neither a specialized enumeration method nor a theorem bounding C-char, so the search behind Table 6.2 is not described. The earlier N<=7 results are backed by the stated case counts and the correctness of Theorems 3.3 and 4.3, so the gap is isolated to the N=8 extension. Nevertheless, the abstract's 'identification of the C-characteristic that occur in ... order up to 17' and the sentence 'no hyperfields with higher C-characteristics are found' are exactly the unsupported part. The text also oscillates between restricting to C-char>2 and reporting only C-char=3; the equivalence of these two statements is part of what needs proof.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies finite real hyperfields whose positive cone P is a cyclic group of order N, and develops an algorithm to enumerate them up to isomorphism. The main theoretical reduction is Theorem 3.3: for a real hyperfield with positive cone P, the hyperaddition is determined by the family {A_x}_{x in P}, where A_x = 1+x, and a family of nonempty subsets of P generates a hyperfield exactly when conditions (kr0)-(kr3) hold. For cyclic P, Theorem 4.3 shows that the symmetry condition (kr3) is automatic once the sets A_{a^i} are defined for i=0,...,floor(N/2) and extended by A_{a^k}=a^k A_{a^{N-k}}, so only (kr0), (kr1), (kr2) need to be checked. The paper provides pseudocode for the construction, the verification of the axioms, the computation of the C-characteristic, and the grouping into isomorphism classes via automorphisms of P. The reported computational results give exhaustive counts for N<=7 (hyperfield orders 3 through 15) and, for N=8, counts of isomorphism classes with C-characteristic equal to 3 under a restriction to C-characteristic exceeding 2. The paper also uses a criterion from the authors' earlier work to conclude that many of the constructed hyperfields are not Krasner quotient hyperfields.","tokens_in":21142,"tokens_out":6009,"duration_ms":58606,"significance":"The reduction of finite real hyperfields with cyclic positive cones to tuples of subsets satisfying three explicit conditions is a clean and useful contribution, and the exhaustive classification for N<=7 is a concrete advance over the previously known enumerations. The proof that (kr3) is redundant in the cyclic case is elegant and makes the search space for N<=7 exactly the stated one. The paper ships pseudocode, software, and a Zenodo database, which strengthens reproducibility. The main significance is however conditional on two points: the claim that the axiom set is minimal is asserted without proof, and the N=8 part of the classification is not backed by a described exhaustive procedure. The conclusion that no hyperfields with C-characteristic higher than 3 occur at N=8 is exactly the kind of negative statement that requires a complete search or a proof; at present it is not reproducible from the manuscript.","major_comments":[{"comment":"The N=8 statement that there are exactly 190 isomorphism classes with C-characteristic 3 and no hyperfields with higher C-characteristics is not supported by the described algorithm. Algorithm 14 is the only enumeration procedure given, and for N=8 it would iterate over (2^8-1)^5 candidate tuples, which is about 1.08e12 cases. The text says that for N>7 the analysis is restricted to C-characteristic exceeding 2, but it does not describe a specialized enumeration, a pruning method, or a theorem bounding the C-characteristic. Consequently the abstract's claim of identifying the C-characteristics occurring in hyperfields of order up to 17, and the sentence 'In this case, no hyperfields with higher C-characteristics are found,' are not justified by the material in the paper. The authors should either provide the exhaustive procedure used for N=8 (with enough detail to reproduce it), prove an upper bound on C-characteristic, or explicitly state that the N=8 results are partial.","section":"Section 6, Table 6.2 and Abstract"},{"comment":"The minimality claim is a central selling point of the paper ('The algorithm is optimal in the sense that the set of axioms used is minimal'), but Remark 3.4 merely asserts that removing any of (kr0)-(kr3) results in a structure that is no longer a hyperfield. No proof or explicit counterexample family is supplied for any of the four axioms. Since this claim concerns load-bearing advertising of the algorithm, it needs a proof or, at minimum, a concise demonstration for each axiom showing a family satisfying the other three conditions that is not a hyperfield. Note also that in the cyclic setting of Section 4, (kr3) is automatic, so if minimality for the implemented algorithm is intended, a separate statement about the minimality of (kr0), (kr1), (kr2) for cyclic cones is needed.","section":"Remark 3.4"},{"comment":"The displayed computation establishing (kr3) ends with '= a^l A_{a^{k-l}} = y A_{y^{-1}x} = x A_{x y^{-1}}'. The last equality, if read literally, is incorrect: x A_{x y^{-1}} with x=a^k equals a^k A_{a^{k-l}}, not a^l A_{a^{k-l}} in general. The intended chain is presumably '= A_{x y^{-1}} y', which would complete the proof because the multiplication group is abelian. This is a typographical issue in a proof, but it should be fixed so that the proof of the redundancy of (kr3) is unambiguous.","section":"Theorem 4.3, proof"}],"minor_comments":[{"comment":"In Algorithm 5, the loop variable in 'for all k in Sum(i,l)' shadows the input parameter k, and the set being accumulated is not clearly connected to the positive part of a^k - a^l. In Algorithm 6, the loop 'for all l in Sum(i,k)' similarly shadows the input parameter l. Renaming the loop variables would make the pseudocode much easier to check.","section":"Algorithm 5 and Algorithm 6"},{"comment":"The line 'Tab[i][N] <- new Entries(-2N-1-i, false, 0)' appears to contain a notational error: the encoding of a singleton negative subset should presumably involve -(2^i) or a similar bit-mask expression, not the subtraction of an index from a constant. This should be corrected or clarified.","section":"Algorithm 16"},{"comment":"The sentence 'is the inverse mapping to (5)' refers to an equation number that does not exist; the intended reference is likely equation (4) or the preceding display.","section":"Section 5, equation (8)"},{"comment":"The introduction says the complete classification covers hyperfields 'of cardinality up to 13', while the abstract and Table 6.1 report orders up to 15 (N=7). This inconsistency should be resolved. In addition, the transition from Table 6.1 (which includes C-char 1 and 2 counts) to Table 6.2 (which lists only C-char 3 for N=8) should explicitly state that C-char 1 and 2 cases for N=8 are excluded by the stated restriction, rather than being absent.","section":"Introduction and Section 6"}],"recommendation":"major_revision","confidential_remarks":"The N=8 claim is the main risk to the paper's advertised scope. The authors appear to have working code and a Zenodo deposit; if that code contains a genuine exhaustive search for N=8 (for example, with pruning based on C-characteristic), then the paper needs only to document that procedure or provide a certificate. If no such exhaustive search exists, I would recommend removing the N=8 rows and softening the abstract accordingly. The minimality claim in Remark 3.4 should also be supported before it is used as a headline contribution."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things you should know upfront. The N<=7 classification of finite real hyperfields with cyclic positive cones looks solid and is a real contribution: Theorem 3.3 gives a clean reduction to the family {A_x} with three conditions, Theorem 4.3 makes (kr3) automatic, and Algorithm 14 is genuinely exhaustive in that range; the case count 260,144,641 for N=7 is exactly (2^7-1)^4, so the brute force is real. The N=8 section, however, is not reproducible from what is written. Section 6 restricts the search to C-characteristic >2 and reports 190 classes with C-char=3 and no higher values, but no specialized search or proof of exhaustiveness is provided, and Algorithm 14 would require roughly 1.08e12 tuples for N=8. That means the 'order up to 17' claim in the abstract currently rests on an undescribed computation.\n\nWhat is new and good: the reduction to half of the A_x sets via Corollary 4.2, the three-axiom check, the large set of non-Krasner examples using Theorem 2.20 (which is a legitimate proved criterion from their earlier paper, not a fitted parameter), and the Zenodo data. The data ship is real evidence, and the N<=7 tables match and extend the prior enumeration literature.\n\nSoft spots, in decreasing severity. First, the N=8 C-characteristic enumeration is unsupported as written: either describe the optimized method, provide code that can be rerun, or prove a bound on C-char for N=8. Second, Remark 3.4 claims the axioms are minimal without proof, and the abstract/title lean on that. Minimality is exactly the kind of thing that needs a short argument or an explicit 'we conjecture'. Third, the scope language is inconsistent: the introduction says 'complete classification ... cardinality up to 13', the abstract says 'order up to 15' and 'order up to 17', and the actual complete classification stops at positive cone order N=7 (hyperfield order 15). That reads as a leftover edit and should be fixed.\n\nOverall, the core framework and the N<=7 enumeration are a genuine step beyond the order-7 classifications in the literature. The flaws are local and repairable. I'd send this to peer review and ask the authors to sort out the N=8 claims and the minimality assertion.","headline":"Solid N<=7 classification of real hyperfields with cyclic positive cones; the N=8 C-characteristic claims are not backed by the described algorithm.","tokens_in":21667,"tokens_out":3504,"would_cite":true,"duration_ms":30914,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["16Y20","12K99","12E20","20N20","08A05"],"pacs":[],"model":"deepseek-v4-flash","headline":"A finite real hyperfield with cyclic positive cone is determined by half the cone data plus three minimal axioms.","keywords":["hyperfield","real hyperfield","cyclic positive cone","C-characteristic","minimal axioms","enumeration algorithm","classification of hyperfields","Krasner quotient hyperfield"],"falsifier":"Run an independent exhaustive enumeration of the defining tuples for $N=8$ with symmetry breaking: if it produces an isomorphism class with C-characteristic 4 or higher, or a count different from 190 for C-characteristic 3, the completeness claim for hyperfield cardinality 17 is refuted.","tokens_in":20703,"feed_emoji":"🧮","tokens_out":12239,"duration_ms":107169,"temperature":0.7,"pith_summary":"The paper establishes that every finite real hyperfield whose positive cone is cyclic is encoded by a short list of subsets: the sets $1+a^i$ for $i=0,\\dots,\\lfloor N/2\\rfloor$, where $N$ is the order of the positive cone. It proves that three conditions on those subsets—covering the cone, a commutativity identity, and an agreement of positive parts—are sufficient to build a real hyperfield, and that removing any one of them lets the structure cease to be a hyperfield. This reduction turns the classification into a finite search over bit-mask tuples, and the paper carries it out completely for positive-cone orders $1,\\dots,7$ (hyperfield cardinalities $3,5,\\dots,15$) and computes the C-characteristic for order $8$ as well. Because every finite real hyperfield whose C-characteristic exceeds 1 is not a Krasner quotient hyperfield, the enumeration yields many new examples outside Krasner's quotient construction.","feed_headline":"Three axioms generate all cyclic-cone real hyperfields","feed_subtitle":"A compact algorithm classifies them up to size 15 and finds many non-Krasner examples.","key_machinery":"The carrying object is the generating family $\\mathcal P=\\{A_x\\}_{x\\in P}$ with $A_x=1+x$, together with the symmetry $A_{a^k}=a^k A_{a^{N-k}}$ that halves the amount of data to $\\lfloor N/2\\rfloor+1$ subsets. Subsets are encoded as exponent bit masks, so each candidate hyperstructure is a tuple of integers; the shift operator implements multiplication by $a^k$, formula (6) extends the family to negative elements, and Algorithm 14 exhausts all tuples while checking (kr0)–(kr2). The three axioms are exactly the conditions, under the cyclic specialization, under which the $A$-family produces a hyperfield.","core_discovery":"The central claim, stated in Theorem 3.3 and Theorem 4.3, is that a finite real hyperfield with cyclic positive cone $P=\\langle a\\rangle$ of order $N$ is completely determined by the family of subsets $A_{a^i}=1+a^i\\subseteq P$ for $0\\le i\\le \\lfloor N/2\\rfloor$, with the remaining sets forced by $A_{a^k}=a^k A_{a^{N-k}}$. Conversely, any such family satisfying (kr0) $\\bigcup_{x\\in P}A_x=P$, (kr1) $A_x+y=A_y+x$, and (kr2) $(A_x-y)\\cap P=(A_{-y}+x)\\cap P$ yields a real hyperfield, and the paper claims these axioms are minimal. The algorithm encodes each candidate by a $(\\lfloor N/2\\rfloor+1)$-tuple of subset codes, verifies the three conditions, computes the C-characteristic, and records one representative per isomorphism class. The reported tables give complete classifications up to hyperfield cardinality 15 and all C-characteristics occurring up to cardinality 17: C-characteristic 2 first appears for cone order 3, C-characteristic 3 first for cone order 7, and the order-8 search finds 190 isomorphism classes with C-characteristic 3 and none higher.","pith_inferences":["The same halving symmetry could be attempted for non-cyclic finite positive cones, where no analogue of the identity $A_{a^k}=a^kA_{a^{N-k}}$ is automatic; finding such a symmetry would be a natural extension of the method.","The reported completeness for order 8 would be strengthened by an independent symmetry-reduced exhaustive search, since the raw tuple space of roughly $(2^8-1)^5$ candidates is too large for the brute-force loop as written.","If the observed pattern that C-characteristic stays bounded by 3 for these orders persists, one could conjecture a bound on C-characteristic in terms of the positive-cone order; the paper does not claim such a bound.","Because the paper only applies the C-characteristic criterion to identify non-quotient examples, one could go further and test each low-order isomorphism class directly against quotient constructions; that would likely sharpen where the quotient/non-quotient boundary lies."],"forward_implications":["The classification tables for cone orders $N=1,\\dots,7$ give explicit counts of hyperfields and isomorphism classes, so the full landscape of these structures is known up to cardinality 15.","Because the axiom set is minimal, any enumeration that uses fewer conditions on the $A_x$ family would admit non-hyperfield structures; the algorithm is optimal within this representation.","Every hyperfield in the enumeration with C-characteristic greater than 1 is, by Theorem 2.20, not a Krasner quotient hyperfield, so the tables supply a large family of new counterexamples to the conjecture that all hyperfields arise by Krasner's construction.","For cone order 8, the search reports exactly 190 isomorphism classes with C-characteristic 3 and no higher C-characteristics, extending the data to hyperfield cardinality 17."],"supporting_citations":[{"why":"Supplies the representation of hyperaddition by the family $A_x=1+x$ and the construction theorem that Theorems 3.2 and 3.3 build on.","marker":"[1]"},{"why":"Supplies Theorem 2.20, the criterion that a finite hyperfield of characteristic 0 with C-characteristic greater than 1 is not a Krasner quotient hyperfield.","marker":"[13]"},{"why":"Introduces Krasner's quotient construction, the class of hyperfields that the paper's new examples are shown not to lie in.","marker":"[15]"},{"why":"Introduces hyperfields and their axioms, the framework against which the minimal axiom set is tested.","marker":"[16]"},{"why":"Provides the first example of a hyperfield that is not a Krasner quotient, the precedent the paper extends to a whole family.","marker":"[22]"},{"why":"Shows that the reversibility axiom (h4) is deducible from the other axioms, supporting the reduction to fewer conditions.","marker":"[23]"},{"why":"Introduces real hyperfields and the notion of a positive cone, the structures the algorithm classifies.","marker":"[26]"}],"fun_headline_variants":["Three minimal axioms generate all cyclic-cone hyperfields","Algorithm classifies cyclic-cone hyperfields to order 15","Non-Krasner hyperfields emerge from minimal axiom algorithm","Finite real hyperfields classified by three axioms"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"For positive-cone order 8, the claim that no C-characteristic above 3 occurs rests on the assumption that the search restricted to C-characteristic greater than 2 found all such hyperfields; the paper gives no proof or optimized procedure that this restriction is exhaustive.","fun_headline_variants_meta":{"raw":{"variants":["Three minimal axioms generate all cyclic-cone hyperfields","Algorithm classifies cyclic-cone hyperfields to order 15","Non-Krasner hyperfields emerge from minimal axiom algorithm","Finite real hyperfields classified by three axioms"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000201,"raw_usage":{"total_tokens":1411,"prompt_tokens":1007,"completion_tokens":404,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":623,"completion_tokens_details":{"reasoning_tokens":339}},"tokens_in":623,"tokens_out":404,"duration_ms":3911,"temperature":1.0,"reasoning_tokens":339,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T16:53:18.144523+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run an independent exhaustive enumeration of the defining tuples for $N=8$ with symmetry breaking: if it produces an isomorphism class with C-characteristic 4 or higher, or a count different from 190 for C-characteristic 3, the completeness claim for hyperfield cardinality 17 is refuted.","supporting_citations":[{"cited_title":"Ameri, M, Eyv azi, S","cited_arxiv_id":null,"evidence_quote":"Supplies the representation of hyperaddition by the family $A_x=1+x$ and the construction theorem that Theorems 3.2 and 3.3 build on."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies Theorem 2.20, the criterion that a finite hyperfield of characteristic 0 with C-characteristic greater than 1 is not a Krasner quotient hyperfield."},{"cited_title":"Krasner, A class of hyperrings and hyperfields, International Journal of Mathematics and Mathematical Sciences 6 (1983) 307-311","cited_arxiv_id":null,"evidence_quote":"Introduces Krasner's quotient construction, the class of hyperfields that the paper's new examples are shown not to lie in."},{"cited_title":"Krasner, Approximation des corps valués complets de caractéristiquep ̸= 0 par ceux de caractéristique 0, Colloque d’Algébre Supérieure, Bruxelles, 1956","cited_arxiv_id":null,"evidence_quote":"Introduces hyperfields and their axioms, the framework against which the minimal axiom set is tested."},{"cited_title":"Massouros, Methods of constructing hyperfields, International Journal of Mathematics and Mathemat- ical Sciences 8 (1985) 725-728","cited_arxiv_id":null,"evidence_quote":"Provides the first example of a hyperfield that is not a Krasner quotient, the precedent the paper extends to a whole family."},{"cited_title":"On the borderline of fields and hyperfields, part II -- Enumeration and classification of the hyperfields of order 7","cited_arxiv_id":"2412.11331","evidence_quote":"Shows that the reversibility axiom (h4) is deducible from the other axioms, supporting the reduction to fewer conditions."},{"cited_title":"Marshall, Real reduced multirings and multifields, Journal of Pure and Applied Algebra 205(2) (2006) 452-468 ALGORITHMIC CONSTRUCTION OF REAL HYPERFIELDS FROM MINIMAL AXIOMS 31","cited_arxiv_id":null,"evidence_quote":"Introduces real hyperfields and the notion of a positive cone, the structures the algorithm classifies."}],"review_version":1}