{"id":"9c1c0786-0f61-4c75-8c09-6256eaa62ca7","arxiv_id":"1908.05775","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Normalized polynomial sequences that give positive bases of skein algebras are bounded by the two Chebyshev families, and on the closed torus only Chebyshev type one works.","lead":"The paper proves that any polynomial recipe for building a positive basis of a surface skein algebra must sit between the two Chebyshev polynomial families, and that on a closed torus only the type-one family works. It gives a precise structural result for the positivity program in skein algebras and cluster theory.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Upper-bound proof relies on sketched low-q cancellations in Lemmas 5.3/5.5; small-n verification would close the gap.","rationale":"The reader identified the Section 2.5 reduction to basic surfaces as the weakest assumption. That reduction is actually justified: for a strict embedding iota:Sigma->Sigma', the map iota_* sends each BP(Sigma) element to a distinct BP(Sigma') element, and the coefficient of any BP(Sigma) basis element in a product inside the subalgebra is exactly the corresponding coefficient in the ambient positive expansion, so positivity descends. Thus the reader's stated concern does not land as a fatal flaw. The more serious soft spot is that Theorem 2.6 for Sigma_{1,1} and Sigma_{0,4} rests on computational lemmas (5.3, 5.4, 5.5) whose proofs are only sketched, with the key lowest-q-term extraction depending on cancellations that are not displayed. I checked the n=3 and n=4 cancellations informally and they are consistent, so no error is demonstrated. A concrete small-n recomputation would settle the concern. The verdict should remain ACCEPT with the reader's moderate confidence, since the paper is not machine-checked and the key computations would benefit from independent verification.","tokens_in":10780,"tokens_out":42839,"duration_ms":395324,"concrete_test":"Independently compute T_{3,1}T_{0,1} and T_{4,1}T_{0,1} on Sigma_{1,1} by direct skein resolution, and S_{3,1}S_{0,1} and S_{4,1}S_{0,1} on Sigma_{0,4}; verify that the lowest q-exponent is exactly q^{-n}S_n((1,0)) with nonnegative coefficients at that exponent. Also resolve the k=0,1 base cases of Lemma 5.4 explicitly and confirm the stated formulas.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Theorem 2.6 is the load-bearing half of the sandwich claim, and its proof for the two basic surfaces reduces to identifying the lowest q-degree term in products of twisted Chebyshev elements. For Sigma_{1,1}, Lemma 5.3 gives T_{n,1}T_{0,1} = q^n T_{n,2} + q^{-n}T_{n,0} + (U+q^2+q^{-2})G_n, and the proof then silently cancels q^{-n}T_{n,0} with the q^{-n}S_{n-2,0} contribution from (U+q^2+q^{-2})G_n to obtain q^{-n}S_n((1,0)); this cancellation is not displayed and would fail if any index in G_n (for example the exponent q^{4i-n-2} or the upper limit floor(n/2)) were off by one. For Sigma_{0,4}, Lemma 5.5 is asserted after 'a routine reduction,' and the claimed separation of the q^{-2n} term from g_n and h_n is the entire proof that (P_n) <= (S_n). These lemmas are stated without full proofs, so the central theorem is only as secure as those computations. This is a rigor gap, not a demonstrated error; the surrounding reduction in Section 2.5 is sound, because for a strict embedding the image of BP(Sigma) is contained in BP(Sigma') and positivity descends to the subalgebra.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves two theorems about positive bases of Kauffman bracket skein algebras of surfaces. Theorem 1 shows that if a normalized polynomial sequence (P_n) with integer coefficients is positive on a surface of genus at least one or with at least four punctures over R = Z[q^{±1}], then the sequence is sandwiched between the normalized Chebyshev polynomials of type one, (T-hat_n), and type two, (S_n). Theorem 2 shows that on the closed torus, the only normalized sequence giving a positive basis is (T-hat_n). The proofs are organized by reducing to three basic surfaces: the closed torus, the once-punctured torus, and the four-punctured sphere. The lower bound is proved in Section 4 using explicit skein resolutions, and the upper bound is proved in Section 5 using product formulas for Chebyshev-type elements on the basic surfaces.","tokens_in":11068,"tokens_out":15772,"duration_ms":130873,"significance":"If the results hold, they provide a sharp constraint on the possible positive bases of skein algebras, strengthening the Fock-Goncharov/Thurston positivity program and giving a surprising uniqueness statement for the torus. The paper contains explicit, parameter-free derivations: the lower bound uses a direct resolution of ab into b_1, b_{-1}, and peripheral products, and the upper bounds use stated product lemmas and the Frohman-Gelca formula for the torus. The theorems are concrete and falsifiable, and the main claims are supported by direct computations rather than by fitting parameters. The proofs are largely self-contained, though some computational steps are compressed.","major_comments":[{"comment":"The reduction to the three basic surfaces is stated as 'This can be seen as follows' and then argued by strict embeddings, but the descent of positivity is not proved in detail. For a strict embedding ι: Σ → Σ', one needs that ι_*(B_P(Σ)) ⊆ B_P(Σ') and that the algebra embedding ι_* preserves positive expansions, so that positivity of B_P(Σ') implies positivity of B_P(Σ). The containment holds because ι is injective on multicurves, but the preservation of the twisted basis under the product expansion is implicit. Since Theorem 1 covers all surfaces with genus ≥ 1 or p ≥ 4, this reduction is load-bearing; the authors should state and prove the descent lemma explicitly.","section":"Section 2.5"},{"comment":"In the proof of Lemma 5.3 for the once-punctured torus, the induction step concludes 'the last equality can be directly verified using the expression of G_n'. This hides the crucial cancellation that produces q^{-n}S_n((1,0)) from the combination of q^{-n}T_{n,0} and the contributions of G_n and A_n. Since the upper bound in Theorem 2.6 for Σ_{1,1} relies on identifying the lowest q-degree term as exactly q^{-n}S_n((1,0)), any off-by-one in the indices or exponents of G_n would break the theorem. The authors should display the verification of the induction step, or at least provide an explicit expansion for small n (e.g., n = 2 and n = 3) and a closed-form verification of the G_{n+1} equality.","section":"Lemma 5.3"},{"comment":"The proof of Lemma 5.5 for the four-punctured sphere ends with 'After a routine reduction, the product S_{n,1}S_{0,1} has the desired form.' This is not a proof: the claimed separation of the q^{-2n}S_{n,0} term from the sums g_n and h_n is exactly the content needed to obtain the upper bound (P_n) ≤ (S_n) in Theorem 2.6 for Σ_{0,4}. The authors should provide the full computation, specifying how the terms in g_n and h_n combine, and verify the q-degree bounds. Without this, the upper-bound theorem is only as secure as an unshown calculation.","section":"Lemma 5.5"}],"minor_comments":[{"comment":"In the first sentence, 'the if a sequence' should be 'if a sequence'.","section":"Abstract"},{"comment":"In the torus proof, the sentence 'For n > 2, (n,2)_T is either P_1((n,2)) or P_2((n/2,1))' is correct but terse; a short explanation that this follows from the parity of n and the already-proved fact P_2 = T_2 would improve readability.","section":"Section 5.1"},{"comment":"The reference [FrG] lists 'F. Charles, and R. Gelca' but the correct authors are C. Frohman and R. Gelca; the title and journal are correct, but the author names should be fixed.","section":"References"},{"comment":"In the proof of Lemma 5.2, the phrase 'apply (n-1)/2 Dehn twists along (1,0) to the equation above' should specify the effect of the Dehn twist on the indices; while the action is standard, writing the resulting index shift would remove ambiguity.","section":"Lemma 5.2"}],"recommendation":"major_revision","confidential_remarks":"The main results are likely correct, and the gaps identified are computational and presentational rather than fundamental. The authors should be asked to fill in the details in Lemmas 5.3 and 5.5 and to state the descent lemma in Section 2.5. These are fixable within the scope of the manuscript. The paper is well-suited to the journal's scope in geometric topology and quantum algebra."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague — quick take on 1908.05775. The headline result is legitimate: the torus rigidity (Theorem 2.5/2) is a clean if-and-only-if classification, and the sandwich (T-hat_n) <= Pn <= S_n over Z[q±1] for g>=1 or p>=4 is a real upper bound that I don't think was known. The genus>=1 lower bound was already in Le's IMRN paper; the genuinely new pieces are the genus-0 lower bound and the upper bound, plus the torus theorem.\n\nWhat the paper does well: proofs are mostly direct skein computations rather than black boxes. The torus proof leverages Frohman-Gelca structure constants, and the argument that positivity forces P2 = T-hat_2 is tidy. The reduction to three basic surfaces via strict embeddings is sound: strict embeddings send BP(Sigma) into BP(Sigma'), so positivity descends. No circularity: the only significant self-citation is the lower bound from Le, which is prior work and independent of the new upper-bound claims.\n\nSoft spots: Lemma 5.3's induction checks out. The 'directly verified' identity is just S_m S_1 = S_{m+1}+S_{m-1} plus index bookkeeping. The stress-test concern about cancellations there does not land on reading. Lemma 5.5 (four-punctured sphere) is genuinely compressed: the proof says 'after a routine reduction' and does not display the verification that q^{-2n}S_{n,0} is the unique lowest-q term, nor fully justify why h_n has no lower degree. The claim is plausible and the statement includes the needed degree bound, but this is the one place a referee should ask for a displayed induction or a small-n check. Also the abstract has a typo ('the if a sequence'). Minor.\n\nVerdict: solid paper for specialists in skein and cluster algebras. It deserves a real referee; I'd recommend accept after minor revision, with a request to expand Lemma 5.5. I'd cite it.","headline":"Torus uniqueness is a genuine classification result and the sandwich bounds are new; the proofs hold up except for one compressed section that a referee should ask to expand.","tokens_in":11595,"tokens_out":3683,"would_cite":true,"duration_ms":34291,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["57N10","57M25"],"pacs":[],"model":"deepseek-v4-flash","headline":"On the closed torus, only the type-one Chebyshev sequence gives a positive skein basis; for other surfaces, every positive normalized sequence lies between the two Chebyshev families.","keywords":["Kauffman bracket skein algebra","positive basis","Chebyshev polynomials","skein algebra","normalized polynomial sequence","surface","cluster algebra"],"falsifier":"Take a normalized sequence with $P_2(x)\\neq \\hat T_2(x)$ on the closed torus, for instance $(S_n)$, and expand $P_1((1,0))P_1((0,1))$ in the associated basis $B_P$ using the torus product formula; if any coefficient falls outside $\\mathbb{Z}_+[q^{\\pm1}]$, that sequence is not positive. A single positive sequence on the torus with $P_2\\neq \\hat T_2$ would disprove the uniqueness claim, and the expansion is a direct check.","tokens_in":10604,"feed_emoji":"🧶","tokens_out":11300,"duration_ms":101316,"temperature":0.7,"pith_summary":"Skein algebras, built from curves on a surface modulo a quantum bracket relation, sit at the center of a positivity conjecture for cluster algebras. The paper asks which normalized polynomial sequences generate positive bases, meaning bases whose multiplication coefficients are nonnegative. It proves a sandwich theorem: any positive normalized sequence must lie between the type-one Chebyshev sequence $(\\hat T_n)$ and the type-two Chebyshev sequence $(S_n)$, for surfaces of genus at least 1 or with at least 4 punctures over $\\mathbb{Z}[q^{\\pm1}]$. On the closed torus the conclusion is much sharper: the sequence $(\\hat T_n)$ is the only one that gives a positive basis. A reader should care because this pins down the exact degree of freedom left for positive bases and explains why the seemingly natural $S_n$ basis fails on the torus.","feed_headline":"Only one Chebyshev family gives a positive torus basis","feed_subtitle":"Every positive skein basis is pinned between the type-one and type-two Chebyshev sequences.","key_machinery":"The central object is the twisted basis $B_P(\\Sigma)$, formed by replacing each component of a simple multicurve $\\gamma=\\prod \\gamma_i^{n_i}$ by $P_{n_i}(\\gamma_i)$. The proof uses the partial order $(P_n)\\leq (Q_n)$ on normalized sequences and reduces every surface to one of three model surfaces: the closed torus $\\Sigma_{1,0}$, the once-punctured torus $\\Sigma_{1,1}$, and the four-punctured sphere $\\Sigma_{0,4}$. On these models, explicit multiplication formulas, including the torus product formula and recursive identities such as $T_{n,1}T_{0,1}=q^n T_{n,2}+q^{-n}T_{n,0}+\\cdots$, provide the control needed to compare an arbitrary $P_n$ with $S_n$ and to force equality with $\\hat T_n$ on the torus.","core_discovery":"Theorem 1 states that a normalized sequence $(P_n)$ with integer coefficients that is positive on a surface of genus at least 1 or with at least 4 punctures satisfies $(\\hat T_n)\\leq (P_n)\\leq (S_n)$, where the inequality means each entry is a nonnegative linear combination of earlier Chebyshev entries. Theorem 2 strengthens the torus case: over $\\mathbb{Z}$ or $\\mathbb{Z}[q^{\\pm1}]$, the sequence $(P_n)$ is positive exactly when $(P_n)=(\\hat T_n)$. The proof shows that positivity of the twisted basis forces $P_1(x)=x$ and then forces each expansion coefficient of $P_n$ relative to the Chebyshev basis to lie in the positive cone; the explicit torus product formula turns this into the uniqueness of $(\\hat T_n)$.","pith_inferences":["My inference: the strict-embedding reduction means the three model surfaces are the only places where positivity can fail; any surface containing one of them as a strict subsurface inherits the same bounds, so the search for positive bases can be focused on the three models.","My inference: the torus uniqueness suggests a broader rigidity principle, namely that any positive canonical basis in a skein algebra, even one not twisted by a single polynomial sequence, will have to reproduce the type-one Chebyshev behavior on every embedded torus.","A testable extension would be to check whether the sandwich $(P_n)\\leq (S_n)$ persists for coefficient rings other than $\\mathbb{Z}[q^{\\pm1}]$ or for bases not generated by one polynomial sequence; the paper does not address these cases."],"forward_implications":["The lower bound $(P_n)\\geq (\\hat T_n)$ now holds for all surfaces of genus at least 1 or with at least 4 punctures, extending the earlier genus-at-least-1 result to the genus-zero case.","The upper bound $(P_n)\\leq (S_n)$ applies to the same family of surfaces, so no positive normalized sequence can exceed the type-two Chebyshev basis in the coefficient order.","On the closed torus the classification is complete: $(\\hat T_n)$ is the only normalized sequence producing a positive basis.","The type-two sequence $(S_n)$ is therefore not positive on the torus, and the paper's conjecture that both $(\\hat T_n)$ and $(S_n)$ are positive on surfaces of negative Euler characteristic remains open."],"supporting_citations":[{"why":"Gives the theorem that isotopy classes of simple multicurves form a free basis of the skein module, which underpins the definition of twisted bases.","marker":"[Pr]"},{"why":"Establishes the lower bound for surfaces of genus at least 1 and supplies the recursion formulas used in the lower-bound proof.","marker":"[Le]"},{"why":"Computes the multiplication rule for the torus in the Chebyshev basis, used to prove both positivity of $(\\hat T_n)$ and its uniqueness.","marker":"[FrG]"},{"why":"Introduced the positive-basis conjecture and proved positivity of $(\\hat T_n)$ when $q=1$, the notion of positivity under study.","marker":"[Th]"},{"why":"Defines the Luo product used to parameterize and multiply curves on the four-punctured sphere in the upper-bound proof.","marker":"[Luo]"}],"fun_headline_variants":["Torus skein positivity forces Chebyshev type-one basis","Only Chebyshev type one yields positive torus skein basis","All positive skein bases lie between two Chebyshev sequences","Positive torus basis is unique and it's Chebyshev type one","Sandwich theorem: positive skein bases squeezed by Chebyshevs"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The reduction to the three model surfaces assumes that positivity of a basis on a larger surface descends to every strictly embedded basic subsurface, so that if the model cases fail, the general case fails; this descent is stated rather than proved in detail.","fun_headline_variants_meta":{"raw":{"variants":["Torus skein positivity forces Chebyshev type-one basis","Only Chebyshev type one yields positive torus skein basis","All positive skein bases lie between two Chebyshev sequences","Positive torus basis is unique and it's Chebyshev type one","Sandwich theorem: positive skein bases squeezed by Chebyshevs"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000173,"raw_usage":{"total_tokens":1186,"prompt_tokens":762,"completion_tokens":424,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":378,"completion_tokens_details":{"reasoning_tokens":333}},"tokens_in":378,"tokens_out":424,"duration_ms":4753,"temperature":1.0,"reasoning_tokens":333,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:08:53.361143+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a normalized sequence with $P_2(x)\\neq \\hat T_2(x)$ on the closed torus, for instance $(S_n)$, and expand $P_1((1,0))P_1((0,1))$ in the associated basis $B_P$ using the torus product formula; if any coefficient falls outside $\\mathbb{Z}_+[q^{\\pm1}]$, that sequence is not positive. A single positive sequence on the torus with $P_2\\neq \\hat T_2$ would disprove the uniqueness claim, and the expansion is a direct check.","supporting_citations":[],"review_version":1}