{"id":"eba508aa-097e-4584-9b82-2fe5a1022c3f","arxiv_id":"2607.26130","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Affine-type cluster algebras have a uniform universal geometric basis — the g-vectors plus one imaginary-wall vector — and an affine exchange matrix stays mutation-finite under adding coefficients exactly when its coefficient rows lie in the imaginary wall.","lead":"This paper proves a 2014 conjecture: in every affine-type cluster algebra, the usual g-vector basis together with one specially chosen extra vector forms a universal basis for the mutation-linear structure, regardless of coefficient ring. It also determines exactly when adding coefficient rows to an affine-type exchange matrix preserves finite mutation-type, correcting a small error in the existing surface-based classification.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 5.1's proof uses the wrong imaginary wall: it places coefficient vectors in d^{B^T}_∞ inside cones of F^B, which is false for simple affine examples; the 'if' direction is not proven as written.","rationale":"The reader's weakest_assumption was the dependence on the companion preprint [17]; that is a genuine verification gap. But the more immediate and internal problem is in the proof of Theorem 5.1: the proof uses d^B_∞ where the theorem and preamble require d^{B^T}_∞, and these walls differ as sets. The rank-2 example shows the proof's route through imaginary cones of F^B cannot work for vectors in d^{B^T}_∞. Because this is an internal inconsistency in a central theorem's proof, it is more directly load-bearing than the [17] dependency. The theorem may be true and fixable, and the reader's conditional verdict already reflects such concerns, so I do not change the verdict. However, the paper should not be accepted until the proof of Theorem 5.1 is corrected or expanded with the missing duality step.","tokens_in":11699,"tokens_out":32359,"duration_ms":271658,"concrete_test":"Take B=[[0,2],[-2,0]] and a=(1,-1)∈d^{B^T}_∞. Explicitly compute the η^B mutation orbit of a; one finds it is periodic of length 3, so the theorem's 'if' direction holds in this example. Then verify that a∉d^B_∞ and a is not in any cone of F^B, contradicting the proof's assertion. This settles that the proof as written is invalid; to resolve the general issue, re-prove the 'if' direction of Theorem 5.1 using the duality between η^B and η^{B^T} on the imaginary wall, or supply a different argument that applies to vectors in d^{B^T}_∞.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The theorem asserts admissibility for B is equivalent to lying in d^{B^T}_∞, and the preamble and Proposition 4.6 consistently use d^{B^T}_∞. However, the proof of Theorem 5.1 states 'we need to show that a vector a∈V* is admissible if and only if it is in d^B_∞' and then, in the 'if' direction, places a∈d^B_∞ in an imaginary cone of F^B. For a skew-symmetric affine B, d^B_∞ and d^{B^T}_∞ are generally distinct; for example, if B=[[0,2],[-2,0]], then d^B_∞ is the ray spanned by (-1,1) while d^{B^T}_∞ is spanned by (1,-1). The coefficient vector (1,-1) lies in d^{B^T}_∞ but not in d^B_∞, so the proof's claim that it lies in an imaginary cone of F^B is false. The subsequent finiteness argument uses linearity of η^B on cones of F^B, which does not apply to cones of F^{B^T}. Thus the 'if' direction of Theorem 5.1 is not established by the given proof. This is not a cosmetic superscript slip: the reduction to imaginary cones of F^B cannot be carried out for vectors in d^{B^T}_∞ without an additional duality argument that is absent from the paper.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper applies the characterization of neighboring seeds in affine type obtained in the companion preprint [17] to prove two results. Theorem 4.1 establishes [14, Conjecture 10.15]: for every exchange matrix B of affine type and every allowed coefficient ring R, the g-vectors of cluster variables associated with B^T, together with the single extra vector -1/2 B^T δ_{B^T}, form a positive basis of the mutation-linear structure R^B. Theorem 5.1 characterizes finite-mutation-type extensions of affine-type exchange matrices: an extension is mutation-finite if and only if each coefficient row lies in the imaginary wall d^{B^T}_∞. The proofs reduce the statements to the finite-type-C case by working with neighboring exchange matrices and the imaginary wall.","tokens_in":11889,"tokens_out":16671,"duration_ms":150833,"significance":"If the results hold, this is a significant contribution. Theorem 4.1 extends the existence of universal geometric coefficients from finite type and low-rank examples to all affine types, uniformly including twisted affine types, and explicitly identifies the extra basis vector conjectured in [14]. Theorem 5.1 gives a uniform, matrix-level criterion for mutation-finiteness with coefficients in affine type and corrects the affine surface/orbifold statements of [6]. The paper is clearly organized and is honest about its dependence on [17]. The proof of Theorem 4.1 is largely coherent and gives an insightful reduction to finite type C via the imaginary wall. The main obstacle to acceptance is that the proof of Theorem 5.1 as written proves a statement with the wrong wall; this is a load-bearing error, though it appears to be repairable by a systematic transpose correction.","major_comments":[{"comment":"The theorem asserts admissibility is equivalent to membership in d^{B^T}_∞, and the discussion preceding the proof states this with the transpose. However, the proof says 'we need to show that a vector a ∈ V* is admissible if and only if it is in d^B_∞' and then, in the 'if' direction, places a ∈ d^B_∞ in an imaginary cone C of F^B. These are not the same wall. For example, for the acyclic affine matrix B = [[0,2],[-2,0]], the wall d^B_∞ is the ray spanned by (-1,1) while d^{B^T}_∞ is spanned by (1,-1), and the latter is not contained in d^B_∞. Consequently the finiteness argument using linearity of η^B_k on cones of F^B does not apply to the vectors required by the theorem. The proof must be rewritten throughout with d^{B^T}_∞, using Propositions 3.2–3.6 with B and B^T interchanged and explicitly verifying that η^B_k sends d^{B^T}_∞ to d^{µ_k(B)^T}_∞.","section":"Section 5, proof of Theorem 5.1"},{"comment":"The main reduction mechanism—Theorem 3.1 and Propositions 3.2–3.6—is quoted from the authors' companion preprint [17], which is a 2025 arXiv preprint whose proofs are not reproduced here. These results are load-bearing for both Theorem 4.1 and the 'if' direction of Theorem 5.1: if any of them fails, the reduction to finite type C collapses. I could not verify these statements from the present manuscript alone. Please either include the statements with more detail, provide an outline of the proofs, or confirm the publication status of [17] so that the referee and readers can check the foundation.","section":"Sections 3–5, reliance on [17]"}],"minor_comments":[{"comment":"The notation ar g(B) is used inconsistently. The theorem needs a basis of ar g(B^T), but the proof says 'We first show that ar g(B) is a spanning set' and later 'These shortest vectors are in g(B^T)'. Please clarify that the spanning set is ar g(B^T) and use ar g consistently throughout.","section":"Section 4, proof of Theorem 4.1"},{"comment":"In the finiteness argument, 'the set of imaginary walls d^{B'}_∞ as B' varies' should be the walls for the transposed matrices, i.e. d^{B'^T}_∞, to match the statement. The authors should also explicitly say that the set of such walls is finite because affine-type exchange matrices form a finite mutation class.","section":"Section 5, proof of Theorem 5.1"},{"comment":"There are several typographical errors: 'MUT A TION-TYPE' in the title line, 'Proposiion' in the labels before Propositions 3.3 and 3.5, and 'The following propositions is' in Section 3. These should be corrected.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The central issue is the wall mismatch in the proof of Theorem 5.1. This is not merely cosmetic: the proof as written establishes the 'if' direction for d^B_∞, not for d^{B^T}_∞, and the two walls differ even in simple rank-2 affine examples. However, the intended proof is clear and likely correct after a systematic transpose correction. The dependence on [17] is also a concern for a journal referee, though it is a legitimate use of a companion preprint. If the authors fix the proof of Theorem 5.1 and address the [17] dependency, the paper should be reconsidered favorably."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's what I'd want you to know before reading: the paper proves a real conjecture (Theorem 4.1) — universal geometric coefficients for affine type, open since 2014 — and gives a uniform mutation-finiteness characterization for affine type with a correction to Felikson-Tumarkin. But the proof of that second theorem (5.1) has a load-bearing transpose error that is not cosmetic, and the if-direction is not proved as written.\n\nWhat's actually new: Theorem 4.1 shows the g-vectors for B^T together with -1/2 B^T δ_{B^T} form a positive basis for the mutation-linear structure R^B. The proof uses the neighboring-seed machinery from the authors' companion preprint [17] to reduce independence to finite type C. That part hangs together, assuming [17] is correct. The spanning argument and the use of Proposition 4.7 are coherent. The result is a genuine advance.\n\nThe soft spot is Theorem 5.1. The statement is about d^{B^T}_∞, but the proof opens \"we need to show... if and only if it is in d^B_∞\" and then, for the if-direction, puts a∈d^B_∞ in an imaginary cone of F^B. For a simple skew-symmetric affine B, say [[0,2],[-2,0]], the two walls are opposite rays: d^B_∞ is spanned by (-1,1) and d^{B^T}_∞ by (1,-1). A vector in d^{B^T}_∞ need not lie in any cone of F^B, so the linearity argument for η^B_k does not apply. Changing the superscripts to d^{B^T}_∞ and F^{B^T} doesn't rescue it, because η^B_k is not linear on F^{B^T}. The Felikson-Tumarkin classification may still imply the theorem, and the paper says as much, but the promised direct proof is not there. Remark 5.2 also says 'twice-punctured annulus' where it should say 'disk' — a smaller inconsistency but in the same region.\n\nThe other caveat is structural: the key propositions (3.1–3.6) are quoted from a 2025 preprint [17] by largely the same authors, with proofs not reproduced. That's not circular, but it makes both theorems conditional on an unreviewed companion. The paper is explicit about this, which I credit.\n\nWho's this for? Specialists in cluster algebra theory, especially people working on universal coefficients and mutation-finiteness. It deserves a serious referee: the conjecture proof is significant, and the mutation-finiteness characterization is plausible and worth fixing. I'd ask a referee to focus on section 5 and require a corrected argument or a clear reduction to the Felikson-Tumarkin result. Send it to review, but with a note that Theorem 5.1's proof needs work.","headline":"Proves the affine universal-coefficients conjecture, but Theorem 5.1's direct proof uses the wrong imaginary wall and needs a real fix, not a typo fix.","tokens_in":12567,"tokens_out":27637,"would_cite":true,"duration_ms":196758,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["13F60"],"pacs":[],"model":"deepseek-v4-flash","headline":"For every exchange matrix of affine type, the g-vectors of cluster variables plus one explicit extra vector form a positive basis of the mutation-linear structure, and an extended matrix is mutation-finite exactly when its coefficient rows","keywords":["cluster algebras","affine type","g-vectors","universal geometric coefficients","mutation-linear algebra","neighboring seeds","imaginary wall","finite mutation-type"],"falsifier":"Take one of the seven explicit 3×3 neighboring blocks listed in Theorem 3.1 (for example, the matrix with rows (0,1,-1), (-1,0,2), (1,-2,0)), compute its mutation fan and the imaginary wall by the recipe of Propositions 3.3–3.4, and verify that the wall is precisely the Minkowski sum of the finite-type-C fan and the imaginary ray. If any extra cone appears, the reduction used in Theorem 4.1's independence proof fails; alternatively, exhibiting a B-coherent linear relation supported inside the imaginary wall that is nontrivial would directly falsify the theorem.","tokens_in":11400,"feed_emoji":"📐","tokens_out":7937,"duration_ms":68219,"temperature":0.7,"pith_summary":"The paper proves two results about cluster algebras of affine type. First, it confirms a conjecture: for any exchange matrix B of affine type, the g-vectors of cluster variables for B^T, together with the vector -1/2 B^T δ_{B^T}, form a positive basis for the mutation-linear structure R^B over every coefficient ring R. Second, it shows that an extended exchange matrix of affine type is mutation-finite if and only if every coefficient row lies in the imaginary wall, the closure of the boundary of the g-vector fan. Both results hold uniformly across all affine types, including twisted ones. The proofs work by reducing statements about affine type to statements about finite type C via neighboring seeds, whose imaginary wall is a direct sum of a finite-type-C fan and one ray.","feed_headline":"One extra vector completes every affine cluster algebra basis","feed_subtitle":"Neighboring seeds reduce affine type to finite type C, yielding universal coefficients and a finiteness criterion.","key_machinery":"The key machinery is the imaginary wall d^B∞, the closure of the boundary of the g-vector fan, together with the notion of a neighboring seed: a seed with n-2 g-vectors lying in a single imaginary cone. For a neighboring exchange matrix B, the wall is a half-hyperplane contained in the hyperplane orthogonal to the imaginary root, and it decomposes as the Minkowski sum of the finite-type-C mutation fan (of the type-C companion Comp_C(B)) and the imaginary ray spanned by -1/2 B δ_B. Any mutation in a non-affine index lifts to an expanded sequence that acts on the wall as a type-C mutation while fixing the imaginary ray pointwise. This reduction to finite type C is what carries both the basis t","core_discovery":"On the paper's own terms, the central discovery is that in affine type the g-vector fan, which fails to cover one codimension-one cone (the imaginary wall), becomes a complete positive basis of the mutation-linear structure once exactly one vector is added: -1/2 B^T δ_{B^T}. Theorem 4.1 states this as the positivity and spanning of that set for every coefficient ring R, resolving Conjecture 10.15. Theorem 5.1 gives a matching characterization of finite mutation-type with coefficients: an extension of B is mutation-finite if and only if all its coefficient rows lie in the imaginary wall of B^T. The proof is carried out by mutating to a neighboring seed, where the imaginary wall restricts to t","pith_inferences":["Our inference: the explicit extra vector -1/2 B^T δ_{B^T} is likely the first in a family of imaginary basis vectors; for indefinite types, one might need one new vector per imaginary direction, and the neighboring-seed technology could guide how to find them.","Our inference: Theorem 5.1 gives an algorithmic test for mutation-finiteness — compute the imaginary wall of B^T (whose inequalities are explicit from Proposition 3.6) and check whether every coefficient row satisfies them; this could be automated for large affine examples.","Our inference: the type-C reduction suggests that other affine-type problems in cluster theory — dominance regions, theta bases, or wall-crossing phenomena — might be transferred to finite type C by the same neighboring-seed machinery, making type-C the 'base case' for affine-type statements."],"forward_implications":["Universal geometric cluster algebras exist for every exchange matrix of affine type, with the basis explicitly given by g-vectors plus -1/2 B^T δ_{B^T}.","The finite-type phenomenon that g-vectors form a basis extends to affine type with exactly one additional vector, uniformly for untwisted and twisted affine types.","An extended exchange matrix of affine type is mutation-finite exactly when all its coefficient rows lie in the imaginary wall; this yields a uniform classification across affine types.","The affine cases of the Felikson–Tumarkin classification for marked surfaces and orbifolds require a redefinition of peripheral laminations (compatibility with every closed curve); the twice-punctured disk is the unique surface where the two definitions differ."],"fun_headline_variants":["Imaginary wall crossed: one vector finishes affine cluster bases","Neighboring seeds yield universal coefficients and finite-type criterion","Finite mutation-type iff coefficient rows lie on imaginary wall","Conjecture 10.15 solved: one extra vector completes affine basis"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The central argument assumes that for any affine-type exchange matrix one can mutate to a neighboring seed for which the boundary (imaginary wall) is exactly a finite-type-C fan plus one fixed ray, with mutations behaving like type-C mutations there; if that structural fact fails, the proofs of both theorems collapse.","fun_headline_variants_meta":{"raw":{"variants":["Imaginary wall crossed: one vector finishes affine cluster bases","Neighboring seeds yield universal coefficients and finite-type criterion","Finite mutation-type iff coefficient rows lie on imaginary wall","Conjecture 10.15 solved: one extra vector completes affine basis"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000202,"raw_usage":{"total_tokens":1150,"prompt_tokens":603,"completion_tokens":547,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":347,"completion_tokens_details":{"reasoning_tokens":475}},"tokens_in":347,"tokens_out":547,"duration_ms":5477,"temperature":1.0,"reasoning_tokens":475,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T00:43:56.432223+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take one of the seven explicit 3×3 neighboring blocks listed in Theorem 3.1 (for example, the matrix with rows (0,1,-1), (-1,0,2), (1,-2,0)), compute its mutation fan and the imaginary wall by the recipe of Propositions 3.3–3.4, and verify that the wall is precisely the Minkowski sum of the finite-type-C fan and the imaginary ray. If any extra cone appears, the reduction used in Theorem 4.1's independence proof fails; alternatively, exhibiting a B-coherent linear relation supported inside the imaginary wall that is nontrivial would directly falsify the theorem.","supporting_citations":[],"review_version":1}