{"id":"bb5f010b-1490-4f62-bfab-ac42ab305e59","arxiv_id":"2411.15079","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"All fake weighted projective planes and K*-surfaces of Picard number one with integral K^2 are described explicitly by 24 series of degree matrices and by pairs of adjacent such matrices.","lead":"This paper classifies all quasismooth, rational, projective surfaces with a torus action, Picard number one, and integral canonical self-intersection, using explicit degree matrices built from Markov-type equations. It also shows that every such K*-surface is encoded by a pair of adjacent fake weighted projective planes, the two central fibers of a degeneration.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 6.11 asserts injectivity of the map from adjacent degree-matrix pairs to K*-surfaces but does not prove it; the l1=l2 case is especially under-addressed.","rationale":"The reader's verdict is CONDITIONAL, and I agree that the paper is not ready in its current form. However, the reader's weakest_assumption attributes the main risk to the external parametrization from [5] and [8] not being reproven. In my reading, the more load-bearing gap is internal: Theorem 6.11's injectivity statement is asserted without proof. The classification of fake weighted projective planes in Theorem 1.1 is a substantial and mostly self-contained achievement, and the typo in the exception sets (eta=1 versus eta=3 in Proposition 4.10) is easily corrected once noticed. But the uniqueness half of Theorem 1.3 is the part that makes the parameterization 'complete' and 'explicit'; without a proof that distinct adjacent ordered pairs give distinct K*-surfaces, the central claim is not established. The l1=l2 case is a concrete place where the missing argument could fail, because the ordered-pair convention does not break the symmetry between the two degenerations. A computational check would either expose a counterexample or, more likely, confirm that the statement is true; either way, the paper needs to supply the missing reasoning. Since my concern does not move the verdict away from CONDITIONAL, I mark the verdict as unchanged.","tokens_in":44312,"tokens_out":13229,"duration_ms":118104,"concrete_test":"Inspect the adjacency data of Remarks 7.11-7.17 and Proposition 7.10 for a pair of distinct adjacent degree matrices (Q1,Q2) with l1=l2 and Q1≠Q2 (for example, in T(1,9) vertices with the same z(2)-Gorenstein index). For each such pair, compute the two surfaces X(Q1,Q2) and X(Q2,Q1) using the formulas of Proposition 6.4 and compare invariants such as the local class group orders, Gorenstein indices, or the explicit Cox ring. If any pair yields isomorphic surfaces, injectivity fails as stated; if all tested pairs are non-isomorphic, the test still demonstrates that the missing uniqueness proof is needed, since the current text gives no criterion ruling out coincidences.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central parameterization claim has two halves: every non-toric K*-surface X is isomorphic to some X(Q1,Q2), and distinct ordered adjacent pairs yield non-isomorphic surfaces. The existence half is supported by the citation to [5] and by Construction 6.1. The uniqueness half, however, is dismissed in a single sentence: 'the definition of a non-toric, ordered pair of adjacent degree matrices ensures that distinct pairs define non-isomorphic K*-surfaces.' No argument is given that the pair (Q1,Q2) is an isomorphism invariant of X, nor that X(Q1,Q2) determines (Q1,Q2) uniquely. In particular, when the two Gorenstein indices satisfy l1=l2, both (Q1,Q2) and (Q2,Q1) satisfy the 'ordered' condition of Definition 6.10. The construction in Construction 6.1 then has a symmetry exchanging the two degenerations, and the paper does not explain why these two ordered pairs should not produce isomorphic K*-surfaces. This is not a minor typo: it is a missing proof of bijectivity in the main theorem. The reader's identified typo in Proposition 4.10 is real but local; the injectivity gap is structural.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper gives an explicit classification of quasismooth, rational, projective surfaces of Picard number one that admit a nontrivial torus action and have integral canonical self-intersection number K^2. The toric case is handled first: using squared Markov-type equations, the authors list all fake weighted projective planes of integral degree as 24 infinite series of degree matrices (Theorem 1.1), compute singularities and T-singularity constellations (Section 5), and then use these data to classify the non-toric K*-surfaces. The main structural result (Theorem 1.3) asserts that every non-toric surface in the class is isomorphic to a surface X(Q1,Q2) attached to a non-toric, ordered pair of adjacent degree matrices, and that distinct ordered pairs give non-isomorphic surfaces. The proof proceeds via degenerations of a K*-surface to two fake weighted projective planes of the same degree, with adjacency formalized in Definitions 6.7, 6.8 and 6.10.","tokens_in":44520,"tokens_out":12664,"duration_ms":107106,"significance":"If the main theorem is correct, this is a complete and explicit parameterization of an interesting class of rational del Pezzo surfaces with torus action, reducing the classification to solutions of squared Markov-type equations and an adjacency relation on degree matrices. The treatment of the Markov equations in Section 2 is elementary and self-contained, and the classification of fake weighted projective planes is carried out with explicit tables and fairly complete proofs. The paper also provides concrete examples, generator matrices, and adjacency graphs (Section 7), which substantially increases its usefulness. The main structural claim, however, relies on an injectivity assertion in Theorem 6.11 that is not actually proved, and there is a concrete inconsistency between Proposition 4.10 and Theorem 1.1 in the degree-1 exception list. Both issues are localizable and in principle fixable, but they affect statements that are load-bearing for the central classification.","major_comments":[{"comment":"The exception list in Proposition 4.10 for degree 1 contains the pair {[1 1 2; 0 1 1], [1 1 2; 0 1 7]}, i.e., it declares (1-8-1) and (1-8-7) to be isomorphic. Theorem 1.1, however, lists the exceptional pair as {[1 1 2; 0 1 3], [1 1 2; 0 1 7]}, i.e., (1-8-3) and (1-8-7). Example 7.7 confirms Theorem 1.1: the generator matrices P2 (η=3) and P4 (η=7) are shown to define isomorphic fake weighted projective planes, while Proposition 5.9 assigns different singularity constellations to (1-8-1) and (1-8-7), so those two cannot be isomorphic. Since Proposition 4.10 is the degree-1 case in the proof of Theorem 1.1, this is not a harmless typo in a remark; the proposition must be corrected, and all later statements that cite it should be rechecked for consistency.","section":"§4, Proposition 4.10; Theorem 1.1; Example 7.7"},{"comment":"The injectivity half of Theorem 6.11 is asserted but not proved. The final sentence of the proof says that 'the definition of a non-toric, ordered pair of adjacent degree matrices ensures that distinct pairs define non-isomorphic K*-surfaces', but no isomorphism invariant of the surface X(Q1,Q2) is shown to recover the ordered pair (Q1,Q2). In particular, when the two local Gorenstein indices satisfy l1=l2, Definition 6.10 allows both (Q1,Q2) and (Q2,Q1) as ordered pairs, and Construction 6.1 has a symmetry that exchanges the two degenerations. The paper does not explain why these two ordered pairs cannot lead to isomorphic K*-surfaces, nor does it provide any argument that the map from ordered pairs to isomorphism classes is well defined and injective. Since the bijectivity statement is precisely the content of Theorem 1.3, this gap is load-bearing and needs a real proof or a reformulation of the parametrization.","section":"§6, Theorem 6.11 and Definition 6.10"}],"minor_comments":[{"comment":"In the definition of an adjusted degree matrix, the clause 'for a ≤ 4' is followed by conditions that refer to a = 9 and a = 8, and the list of allowed η values appears incomplete (for example, the degree-1 series (1-8-7) requires η=7, which is not listed). Please correct the case distinction and the displayed values.","section":"§3, Definition 3.14"},{"comment":"The displayed matrix in Case (ii) is written as [x0^2 x1^2 2x2^2; η̄0 η̄1 η̄2] with the accompanying text '¯0, ¯1, ¯η ∈ Z/8Z', which is garbled; it should presumably say that the second row is (0̄, 1̄, η̄) with η̄ ∈ {1̄,3̄,5̄,7̄}.","section":"§4, proof of Proposition 4.10, Case (ii)"},{"comment":"In the list of self-adjacent cases, the block labeled 'u = (1, 2, 3)' appears twice; the third block should correspond to u = (1, 4, 5) for the degree-5 series (5-1-0), (1-5-1), (1-5-4).","section":"§7, Remark 7.18"}],"recommendation":"major_revision","confidential_remarks":"The overall approach is plausible and the manuscript contains much valuable explicit computation, but the two major issues—the inconsistent exception list in Proposition 4.10 and the unproved injectivity in Theorem 6.11—need to be fixed before the classification claim can be accepted. The injectivity gap is the more serious one; it affects the central bijection and may require either a genuine invariant or a modification of the statement for the l1=l2 case. I would advise the editor to request a revised version addressing these points."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This paper gives an explicit classification of quasismooth rational projective surfaces of Picard number one with torus action and integral K^2. The toric half is the genuine advance: 24 series of degree matrices for fake weighted projective planes of integral degree, derived from a clean self-contained treatment of squared Markov equations. The singularity analysis in Section 5 (local Gorenstein indices, T-singularities) is useful and complements Hacking–Prokhorov. The adjacency-pair reformulation of K*-surfaces is a nice framing.\n\nThe reader is right that Proposition 4.10 contains a local error: the exception set in degree 1, series (1-8-*), lists {[1 1 2; 0 1 1], [1 1 2; 0 1 7]}, while Theorem 1.1 and Example 7.7 both point to {[1 1 2; 0 1 3], [1 1 2; 0 1 7]}. This is a typo, but it makes Theorem 1.1 as written inconsistent with its proof.\n\nThe larger soft spot is in the uniqueness half of Theorem 6.11 (and Theorem 1.3). The proof dispatches injectivity with one sentence: 'the definition ... ensures that distinct pairs define non-isomorphic K*-surfaces.' No argument is given that the ordered pair (Q1,Q2) is an isomorphism invariant of X, nor is the l1=l2 case addressed. When the two local Gorenstein indices coincide, both (Q1,Q2) and (Q2,Q1) satisfy the ordered condition, and the paper does not say whether these produce isomorphic surfaces or whether the pairs are identified. As written, the classification statement includes a claim that is not established. This is a structural gap, not a fixed typo.\n\nThe existence half of the theorem is supported by the earlier parametrization from [5] and [8]; that dependency is legitimate but should be flagged explicitly. The Markov/toric part is derived from scratch, and the tables give enough data to check many of the isomorphy claims.\n\nFor a reader working on torus actions or del Pezzo surfaces with quotient singularities, this is a genuinely useful paper, once the injectivity issue is settled. It deserves a serious referee—the toric classification alone merits it—and the review should ask the authors to prove, or fix the statement of, the injectivity claim and correct the Proposition 4.10 typo.","headline":"A substantial, largely self-contained toric classification with a real but fixable gap in the K*-surface parametrization: injectivity of ordered adjacent pairs is asserted, not proved.","tokens_in":45090,"tokens_out":4854,"would_cite":false,"duration_ms":44757,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14L30","14J26"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper's central claim is that every quasismooth, rational, projective surface of Picard number one with a nontrivial torus action and integral canonical self-intersection number is explicitly classified by ordered pairs of adjacent…","keywords":["K*-surfaces","Picard number one","fake weighted projective planes","squared Markov type equations","degree matrices","toric degenerations","T-singularities","integral canonical degree"],"falsifier":"A concrete way to test the classification is to enumerate the first few levels of the Markov mutation trees for $a=9,8,6,5$, compute the adjacency graph of the corresponding fake weighted projective planes with the formulas of Proposition 6.4, and compare with the graphs drawn in Remarks 7.12–7.17; an extra or missing edge at any level would refute completeness of the adjacency description. Alternatively, exhibiting two distinct non-toric ordered adjacent pairs $(Q_1,Q_2)$ and $(Q_1',Q_2')$ whose associated $K^*$-surfaces $X(Q_1,Q_2)$ and $X(Q_1',Q_2')$ are isomorphic would refute Theorem 1.3 directly.","tokens_in":44080,"feed_emoji":"🧮","tokens_out":11503,"duration_ms":102937,"temperature":0.7,"pith_summary":"A complete explicit classification is given for quasismooth, rational, projective surfaces of Picard number one that admit a nontrivial torus action and have integral canonical self-intersection number $K_X^2$. In the toric case these are the fake weighted projective planes of integral degree, classified by 24 infinite series of degree matrices built from solutions of the squared Markov-type equation $(u_0+u_1+u_2)^2 = a\\,u_0u_1u_2$. In the one-dimensional torus case, every non-toric $K^*$-surface in the class is shown to be isomorphic to the surface $X(Q_1,Q_2)$ attached to an ordered pair of adjacent degree matrices, and distinct ordered pairs give non-isomorphic surfaces. The possible degrees are forced to lie in $\\{1,2,3,4,5,6,8,9\\}$, so the classification has finitely many series but infinitely many members generated by Markov-type mutations.","feed_headline":"24 series capture every integral-degree torus-action surface","feed_subtitle":"Each is encoded by two adjacent degree matrices built from squared Markov-type equation solutions.","key_machinery":"The central computational object is the degree matrix $Q=[q_0,q_1,q_2]$ with columns $q_i=(u_i,\\bar\\eta_i)$ in the divisor class group $\\mathbb{Z}\\oplus\\mathbb{Z}/\\mu\\mathbb{Z}$; the first row $(u_0,u_1,u_2)$ is a solution of the squared Markov-type equation $(u_0+u_1+u_2)^2=a\\,u_0u_1u_2$, and the columns encode the Cox ring generators. Two degree matrices are called adjacent when they are the central fibers of one equivariant flat family of $K^*$-surfaces (Construction 6.1); Proposition 6.4 recasts adjacency as a concrete condition involving a T-singularity (a cyclic quotient singularity of type $\\frac{1}{dk^2}(1,dpk-1)$) and a local Gorenstein index of one toric fixed point. Adjusted degree matrices (Definition 3.14) make the encoding unique up to isomorphism, so ordered adjacent pairs $(Q_1,Q_2)$ parametrize the $K^*$-surfaces $X(Q_1,Q_2)$ without duplication. The mutation operation $\\lambda$ on solution triples drives the infinite series in the tables.","core_discovery":"On its own terms, the paper's discovery is a complete classification statement for the class in the title. Its Theorem 1.1 says that a fake weighted projective plane of integral degree is, up to isomorphism, exactly one of the surfaces $Z(Q)$ attached to a degree matrix in the 24-series table, with distinct matrices giving non-isomorphic planes except in three explicitly listed exceptional triples. Its Theorems 6.9 and 6.11 upgrade this to $K^*$-surfaces: a non-toric, quasismooth, rational, projective $K^*$-surface of Picard number one with integral $K_X^2$ is isomorphic to a unique $X(Q_1,Q_2)$, where $(Q_1,Q_2)$ is a non-toric ordered pair of adjacent degree matrices. Here adjacency is a geometric relation: two fake weighted projective planes are adjacent exactly when they are the two central fibers of the flat degenerations of one common $K^*$-surface, so the classification is not just a list of isolated surfaces but a description of the degeneration graph on the toric side.","pith_inferences":["Because the mutation tree of the squared Markov-type equation is recursively generated, the classification yields an algorithm for enumerating these $K^*$-surfaces level by level; the paper itself draws the first levels of these graphs for several series but does not state the enumeration as a theorem.","The appearance of mutations of triples and adjacent degree matrices suggests a cluster-algebra interpretation: adjacency may be the geometric shadow of matrix mutation, and a cluster structure could organize the countably infinite parameterization into finite-rank seeds.","The explicit list gives a direct test bed for questions outside this paper, such as which of these surfaces carry Kähler–Einstein metrics or are smoothable; the T-singularity data in Section 5 is the kind of input those criteria require.","The paper stops at integral $K_X^2$; if the squared Markov-type equation is replaced by a non-square parameter, the same degenerations and adjacency machinery would still produce a candidate classification of the non-integral case, though the table of series would grow or change."],"forward_implications":["Every non-toric quasismooth rational projective $K^*$-surface of Picard number one with integral $K_X^2$ is now explicitly parametrized; the parameter space is the countable set of non-toric ordered adjacent pairs of degree matrices from the 24-series table.","The possible values of $K_X^2$ are confined to $\\{1,2,3,4,5,6,8,9\\}$; in particular no such surface has $K_X^2=7$.","For fake weighted projective planes of integral degree, the classification pinpoints the T-singularity locus: exactly the seven series $(2\\text{-}3\\text{-}1)$, $(1\\text{-}8\\text{-}1)$, $(1\\text{-}8\\text{-}5)$, $(1\\text{-}6\\text{-}1)$, $(1\\text{-}5\\text{-}1)$, $(1\\text{-}5\\text{-}2)$, $(1\\text{-}5\\text{-}3)$ have a unique T-singularity among three singularities, and all other series have at most T-","The adjacency relation turns each family of fake weighted projective planes of fixed degree and multiplicity into a graph; for degrees $9,8,6,5,4,3$ these graphs are connected and often isomorphic to the Markov tree $T(a)$ of ascending solution triples.","Distinct ordered adjacent pairs yield non-isomorphic $K^*$-surfaces, so the parametrization is faithful, not only surjective."],"supporting_citations":[{"why":"It supplies the Cox-ring and degree-matrix formalism, including the isomorphism criterion for fake weighted projective spaces used throughout the classification.","marker":"[2]"},{"why":"It provides the theory of T-singularities and the prior classification of T-singular projective toric surfaces of Picard number one that the singularity tables complement.","marker":"[4]"},{"why":"It establishes the parametrization of non-toric del Pezzo surfaces of Picard number one with torus action by the family $X(P)$, which is the starting point of the $K^*$-surface classification.","marker":"[5]"},{"why":"It gives the formulas for canonical self-intersection numbers and local Gorenstein indices of toric varieties that the proofs of the degree and singularity tables repeatedly use.","marker":"[6]"},{"why":"It contains the degeneration construction and the equality of canonical degrees across the flat families, the mechanism by which pairs of adjacent degree matrices arise from a $K^*$-surface.","marker":"[8]"},{"why":"It is the classical source for Markov-type equations and their mutations, which underpin Theorem 2.2 and the solution triples $S(a)$.","marker":"[10]"}],"fun_headline_variants":["24-series table completes all integral K*-surfaces","Adjacent matrices classify every integral torus surface","Every integral-degree torus surface fits a 24-series","Complete map: torus surfaces with integral self-intersection"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument leans on a previously established parametrization, namely that every non-toric, quasismooth, rational, projective $K^*$-surface of Picard number one is isomorphic to one of the explicit surfaces $X(P)$ written with a $3\\times 4$ generator matrix in Section 6; if that parametrization misses a surface, the new classification would be incomplete.","fun_headline_variants_meta":{"raw":{"variants":["24-series table completes all integral K*-surfaces","Adjacent matrices classify every integral torus surface","Every integral-degree torus surface fits a 24-series","Complete map: torus surfaces with integral self-intersection"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000762,"raw_usage":{"total_tokens":3293,"prompt_tokens":769,"completion_tokens":2524,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":385,"completion_tokens_details":{"reasoning_tokens":2468}},"tokens_in":385,"tokens_out":2524,"duration_ms":19163,"temperature":1.0,"reasoning_tokens":2468,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T14:31:22.183047+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete way to test the classification is to enumerate the first few levels of the Markov mutation trees for $a=9,8,6,5$, compute the adjacency graph of the corresponding fake weighted projective planes with the formulas of Proposition 6.4, and compare with the graphs drawn in Remarks 7.12–7.17; an extra or missing edge at any level would refute completeness of the adjacency description. Alternatively, exhibiting two distinct non-toric ordered adjacent pairs $(Q_1,Q_2)$ and $(Q_1',Q_2')$ whose associated $K^*$-surfaces $X(Q_1,Q_2)$ and $X(Q_1',Q_2')$ are isomorphic would refute Theorem 1.3 directly.","supporting_citations":[{"cited_title":"144, Cambridg e University Press, Cambridge,","cited_arxiv_id":null,"evidence_quote":"It supplies the Cox-ring and degree-matrix formalism, including the isomorphism criterion for fake weighted projective spaces used throughout the classification."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It provides the theory of T-singularities and the prior classification of T-singular projective toric surfaces of Picard number one that the singularity tables complement."},{"cited_title":"Del Pezzo surfaces of Picard number one admitting a torus action","cited_arxiv_id":"2207.14790","evidence_quote":"It establishes the parametrization of non-toric del Pezzo surfaces of Picard number one with torus action by the family $X(P)$, which is the starting point of the $K^*$-surface classification."},{"cited_title":"Classifying log del Pezzo surfaces with torus action","cited_arxiv_id":"2302.03095","evidence_quote":"It gives the formulas for canonical self-intersection numbers and local Gorenstein indices of toric varieties that the proofs of the degree and singularity tables repeatedly use."},{"cited_title":"On Degenerations of the Projective Plane","cited_arxiv_id":"2405.04862","evidence_quote":"It contains the degeneration construction and the equality of canonical degrees across the flat families, the mechanism by which pairs of adjacent degree matrices arise from a $K^*$-surface."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It is the classical source for Markov-type equations and their mutations, which underpin Theorem 2.2 and the solution triples $S(a)$."}],"review_version":1}