{"id":"52042449-ed4a-4d17-b0fd-dc413d9454ac","arxiv_id":"2411.17579","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For monomial algebraic and algebroid curves over a noetherian ring R, the relative Lipschitz saturation equals A[t^{L(r)}], where L(r) is computed explicitly from the numerical semigroup of the exponents.","lead":"This mathematics paper gives an explicit formula for the Lipschitz saturation of monomial curve algebras over a ring R in terms of the numerical semigroup generated by the exponents. The result makes a previously abstract computation concrete for a broad class of monomial algebraic and algebroid curves, which is useful for singularity theory and commutative algebra.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Proposition 2.5 is false: the valuative criterion gives only integral closure; for α=(5,7), x^8-y^8 is not in (x^5-y^5,x^7-y^7), so Theorem 3.14 fails for Γ=<5,7>.","rationale":"The reader's weakest_assumption focuses on noetherianness, γ1-niceness, Observation 3.1, and (Γ,B)-closedness, but the foundational algebraic lemma is unsound. The proof of Proposition 2.5 applies the Valuative Criterion of Swanson-Huneke, which is an if-and-only-if criterion for membership in the integral closure overline(I). The DVR verification in the paper establishes h ∈ overline(I), not h ∈ I. In general overline(I) is strictly larger than I, and here it is: I = (x^5-y^5, x^7-y^7) in k[x,y] does not contain x^8-y^8, as the homogeneous coefficient system shows. The DVR argument is consistent because x^8-y^8 lies in overline(I); indeed every DVR image lands in IV, which is exactly what the valuative criterion detects. This invalidates Corollary 2.6 and the lower inclusion in Proposition 3.3/Theorem 3.14. Concretely, for Γ=<5,7>, R=k, B=k[t], A=k[t^5,t^7], the theorem's formulas give A* = A[t^8,t^9,t^11], but t^8 fails the required membership Δ(t^8) ∈ ker φ. The central claim is therefore false as stated, so the paper should be rejected unless Proposition 2.5 is replaced by a true statement (for example, an integral-closure statement) and the main theorem is appropriately restricted or reformulated.","tokens_in":15808,"tokens_out":23537,"duration_ms":179113,"concrete_test":"Run an ideal-membership test in a CAS (Macaulay2/Singular): compute the normal form of x^8-y^8 with respect to a Gröbner basis of (x^5-y^5, x^7-y^7) in k[x,y], or solve the degree-3/-1 coefficient system above. If the normal form is nonzero, Proposition 2.5 is false. As a second check, compute the saturation k[t^5,t^7]^*_{k[t],k} directly and verify that t^8 is absent, contradicting Theorem 3.14 for Γ=<5,7>.","verdict_should_be":"REJECT","load_bearing_attack":"The central membership step, Proposition 2.5, is false. Its proof invokes the Valuative Criterion [22, Theorem 6.8.3], which characterizes membership in the integral closure of an ideal, not membership in the ideal itself; the DVR argument establishes at most h ∈ overline(I). In the noetherian ring T = k[x,y], take z1=x, z2=y, α1=5, α2=7, d=1. The proposition claims x^8-y^8 ∈ I := (x^5-y^5, x^7-y^7). A homogeneous degree-8 representation x^8-y^8 = A(x^5-y^5) + B(x^7-y^7) with deg A=3 and deg B=1 forces the coefficient equations a0+b0=1, a1+b1=0, a2=0, a3=0, -a0=0, -a1=0, -a2-b0=0, -a3-b1=-1, which are inconsistent (a0=0 and b0=0 contradict a0+b0=1). Hence x^8-y^8 ∉ I, so Corollary 2.6 is false and t^8 ∉ R[t^5,t^7]^*_{R[t],R}. Consequently Theorem 3.14 fails: for Γ=<5,7>, R=k, B=k[t], A=k[t^5,t^7] satisfies all hypotheses ((Γ,B)-closed by Proposition 3.8(a), R noetherian and γ1-nice), but the theorem predicts t^8 ∈ A* via L(2)={8,9,11}, contradicting the ideal-membership check.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript studies the relative Lipschitz saturation A*_{B,R} of a monomial subalgebra A ⊆ R[[t^{γ_1},...,t^{γ_n}]] inside an R-subalgebra B ⊆ R[[t]] containing R[t]. Its main result, Theorem 3.14, claims an explicit description A*_{B,R} = A[t^{L(r)}] in terms of the numerical semigroup Γ = ⟨γ_1,...,γ_n⟩, under the assumptions that R is a noetherian γ_1-nice ring and that A is (Γ,B)-closed. Section 2 develops algebraic membership results for binomial differences, and Section 3 uses those results, together with Proposition 3.4, to prove the theorem. The appendix lists computed examples for several semigroups.","tokens_in":16075,"tokens_out":13786,"duration_ms":156010,"significance":"If correct, the result would give a useful and explicit computation of relative Lipschitz saturation for monomial curves over general ground rings, extending earlier analytic results. The paper introduces convenient notation involving partial gcds and gap sets, and the reverse-inclusion part of the argument is plausible. However, the central claim is false: the proof of Proposition 2.5 mistakes a valuative criterion for integral closure for a criterion of ideal membership, and a concrete counterexample satisfying all hypotheses of Theorem 3.14 breaks the main result. The paper therefore does not establish its advertised description.","major_comments":[{"comment":"Proposition 2.5 is false. The proof invokes the Valuative Criterion [22, Theorem 6.8.3], which characterizes membership in the integral closure of an ideal, not membership in the ideal itself; the DVR argument establishes at most h ∈ \\overline{I}. In the noetherian ring T = k[x,y] with z1 = x, z2 = y, α1 = 5, α2 = 7, d = 1, the proposition claims x^8 − y^8 ∈ I = (x^5 − y^5, x^7 − y^7). Since I is homogeneous, any degree-8 representation must take the form A(x^5 − y^5) + B(x^7 − y^7) with deg A = 3 and deg B = 1. Writing A = a0 x^3 + a1 x^2 y + a2 x y^2 + a3 y^3 and B = b0 x + b1 y, the coefficients of x^8, x^6 y^2, x^3 y^5, and y^8 give a0 + b0 = 1, a2 = 0, −a0 = 0, and −a3 − b1 = −1, so a0 = 0 and b0 = 0, contradicting a0 + b0 = 1. Hence x^8 − y^8 ∉ I, and Proposition 2.5 is false.","section":"§2, Proposition 2.5"},{"comment":"The failure of Proposition 2.5 invalidates Corollary 2.6 and Proposition 3.3, and consequently Theorem 3.14. For a concrete disproof of the main theorem, take a field k with char k ≠ 5, R = k, B = k[t], A = k[t^5, t^7], and Γ = ⟨5,7⟩. All hypotheses of Theorem 3.14 hold: Γ is coprime, 5 ∤ 7, d2 = 1, R is noetherian and γ1-nice, and A is (Γ,B)-closed by Proposition 3.8(a). Corollary 2.4, which is correct, gives ker φ = ⟨t1^5 − t2^5, t1^7 − t2^7⟩ in k[t1,t2]. If t^8 belonged to A*_{B,R}, then t1^8 − t2^8 would lie in that ideal, but the degree argument in the previous comment shows it does not. Thus t^8 ∉ A*_{B,R}, whereas Theorem 3.14(c) predicts t^8 ∈ A[t^{L(2)}] because 8 ∈ L~(2) ⊆ L(2). The main theorem is therefore false.","section":"§2 Corollary 2.6 and §3 Theorem 3.14"},{"comment":"Observation 3.1 is asserted without proof. It claims that for C ⊆ k[[t]] and A′ ⊆ k[[x1,...,xs]] ∩ C, the kernel of C ⊗_k C → C ⊗_{A′} C is contained in the ideal ⟨Δ(x1),...,Δ(xs)⟩ in k[[t1,t2]]. This requires passing to the completed tensor product, justifying injectivity of C ⊗_k C → k[[t1,t2]], and proving that the image of the kernel is contained in the closed ideal generated by the Δ(xi). None of these steps is given, yet Proposition 3.4 and consequently Lemma 3.12 rely on this inclusion.","section":"§3, Observation 3.1"}],"minor_comments":[{"comment":"The manuscript contains a 'DRAFT - GSN - 2022' header and numerous OCR-type artifacts (e.g., '/shortrightarrow' symbols and garbled ligatures); these should be cleaned before any submission.","section":"Global"},{"comment":"The notation L_j is redefined as L_j := {ℓ ∈ L_j | ℓ ≤ γ_j + γ1 − 1} after L_j was already defined in Section 3; this reuse is confusing and a different symbol, such as L_j^#, would be clearer.","section":"§3, before Lemma 3.12"},{"comment":"The appendix examples derive their claimed equalities from Theorem 3.14; since the theorem is false, these examples should be rechecked independently of the theorem before being presented as illustrations.","section":"Appendix A"},{"comment":"Part (c) writes A*_{B,R} = A[t^{L(r)}] = A[t^{L(r)}], but the two displayed sets are not defined identically (the truncated L(r) versus the full L(r)); the intended relationship should be stated precisely with distinct symbols.","section":"Theorem 3.14(c)"}],"recommendation":"reject","confidential_remarks":"The counterexample is elementary and decisive, and the error in Proposition 2.5 is not a local gap but a misapplication of the valuative criterion. I do not see how the main theorem could be repaired within the manuscript's current framework: the claimed equality between the Lipschitz saturation and the semigroup-generated algebra is false. The authors might prove a related statement about the integral closure of A in B, but that is not the result claimed in the abstract and introduction."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's the bottom line: the central claim does not survive contact with a simple example. Proposition 2.5 tries to show z1^{αm+d}-z2^{αm+d} lies in the ideal generated by the z1^{αi}-z2^{αi}, but the proof only establishes membership in the integral closure, using the valuative criterion. Those are different. Take T=k[x,y], α=(5,7), d=1. Then x^8-y^8 has no homogeneous degree-8 representation in (x^5-y^5,x^7-y^7); the coefficient equations force a0+b0=1 and a0=b0=0. So Corollary 2.6 and Proposition 3.3 are false, and Theorem 3.14 fails for Γ=<5,7> with B=k[t], A=k[t^5,t^7]: the predicted t^8 is not in A^*_{B,R}.\n\nThat is the load-bearing flaw. I want to give credit where it is due: the overall strategy—describing the saturation via the semigroup and partial gcds—is natural, and the relative setting over a general ring is a reasonable extension. The early lemmas (2.1–2.4) are correct, and the appendix examples show the formula in action. The (Γ,B)-closed notion is a sensible condition, though not characterized.\n\nThe secondary issues are real but minor by comparison. Observation 3.1 is stated without proof and relies on passing to the completed tensor product; that needs a careful argument. The appendix examples would also be more convincing with a reproducible script, but they are not the problem.\n\nIf the paper were correct, it would be a useful tool for the Lipschitz-saturation community. As it stands, the main theorem is false, and the error is not a gap you can patch with a remark—it is a counterexample. The authors may be able to recover something by rephrasing the result in terms of integral closure, but that would be a different paper.\n\nFor peer review: send it out if you want, because the mistake is instructive and the topic is legitimate. But the referee report should be a rejection or a demand for a major reworking. I would not cite it in its current form.","headline":"The main theorem is false: Proposition 2.5 confuses integral closure with ideal membership, and Γ=<5,7> gives a concrete counterexample.","tokens_in":16680,"tokens_out":5777,"would_cite":false,"duration_ms":47988,"reading_group":"no","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["13B22","14H20"],"pacs":[],"model":"deepseek-v4-flash","headline":"The Lipschitz saturation of a monomial curve algebra is explicitly determined by the gap structure of its numerical semigroup.","keywords":["Lipschitz saturation","relative Lipschitz saturation","numerical semigroup","monomial curve","algebroid curve","power series ring","bi-Lipschitz equisingularity","integral closure"],"falsifier":"Take $R = \\mathbb{F}_2$, $\\Gamma = \\langle 2, 3 \\rangle$, $B = \\mathbb{F}_2[t]$, and $A = \\mathbb{F}_2[t^2,t^3]$, then compute $A^*_{B,R}$ directly from the definition; here $\\operatorname{char} R$ divides $\\gamma_1$, so the theorem is not asserted, and comparing the result with the predicted $A[t^{L(r)}]$ would show whether the characteristic hypothesis is essential.","tokens_in":15530,"feed_emoji":"🧮","tokens_out":8217,"duration_ms":73165,"temperature":0.7,"pith_summary":"This paper tries to turn the Lipschitz saturation of a monomial curve algebra into a computation rather than an abstract closure operation. For a subalgebra $A = R[t^{\\gamma_1},\\ldots,t^{\\gamma_n}]$ inside a larger ring $B$ of power series in $t$, it claims that the saturation $A^*_{B,R}$ is obtained by adjoining a finite set of monomials $t^\\ell$, with the exponents determined solely by the numerical semigroup $\\Gamma$ generated by the $\\gamma_i$. The description is uniform across three natural choices of $B$: polynomials, all power series, and convergent analytic series over $\\mathbb{R}$ or $\\mathbb{C}$. Because the extra monomials are explicit, the result reduces a question about tensor products and integral closures to a semigroup calculation, which matters since Lipschitz saturation is the algebraic counterpart of bi-Lipschitz equisingularity of curve germs.","feed_headline":"Monomial curve saturations pinned down by semigroup gaps","feed_subtitle":"For monomial curve algebras, Lipschitz saturation is a finite list of extra monomials read off from the numerical semigroup.","key_machinery":"The load-bearing object is the relative Lipschitz saturation $A^*_{B,R}$, defined for $R$-algebras $A \\subseteq B$ through the diagonal map $\\Delta(x) = x \\otimes 1 - 1 \\otimes x$ and the kernel of the canonical map $B \\otimes_R B \\to B \\otimes_A B$: an element of $B$ is saturated when $\\Delta(x)$ lies in that kernel. The proof's engine is a pair of semigroup facts: a divisibility and induction lemma shows that once monomials of degrees $\\alpha$ and $\\alpha+d$ are saturated, every monomial of degree $\\alpha + sd$ is saturated; and a root-of-unity argument shows that in a saturated series no coefficient can sit in a gap degree not divisible by the relevant partial gcd, provided the residue characteristic does not divide $\\gamma_1$. These forces combine with the gap structure of $\\Gamma$ to leave exactly the listed monomials.","core_discovery":"The paper's central claim is Theorem 3.14: under a noetherian $\\gamma_1$-nice ground ring $R$ and the assumption that $A$ is $(\\Gamma,B)$-closed, the Lipschitz saturation equals $A[t^{L(r)}]$, where $r$ is the first index with $d_r = \\gcd(\\gamma_1,\\ldots,\\gamma_r)=1$, and $L(r)$ is the union of the intermediate gap sets $L_j = \\{\\ell \\in G(\\Gamma) : \\gamma_j < \\ell < \\gamma_{j+1},\\ d_j \\mid \\ell,\\ \\ell \\leq \\gamma_j + \\gamma_1 - 1\\}$ together with the terminal interval $\\tilde L(r) = \\{\\ell \\in G(\\Gamma) : \\gamma_r + 1 \\leq \\ell \\leq \\gamma_r + \\gamma_1 - 1\\}$. The same equality has explicit $R$-module and $A$-module versions. Thus the saturation is a finitely generated $A$-algebra with an explicit monomial presentation, and in the two-generator case with $\\gamma_1 = 2$ the saturation collapses to $A$ itself.","pith_inferences":["A likely next step is to characterize $(\\Gamma,B)$-closedness in terms of the conductor of $\\Gamma$ or the coefficient module of $B$, which would extend the theorem beyond the three choices of $B$ checked in the paper.","Because Lipschitz saturation sits between $A$ and its normalization, the same gap-set formula may yield a direct algorithm for the integral closure or normalization of monomial algebroid curves.","The characteristic assumption enters only through the root-of-unity step, so small examples with $\\operatorname{char} R$ dividing $\\gamma_1$ would show whether that hypothesis is essential or merely an artifact of the proof.","The finite set $L(r)$ could be packaged as a new semigroup invariant, analogous to the Apéry set, and compared with other invariants of numerical semigroups."],"forward_implications":["Under the theorem's hypotheses, the Lipschitz saturation of a monomial curve algebra is a finitely generated $A$-algebra with an explicitly listed monomial generating set.","For the polynomial, formal power series, and convergent analytic choices of $B$, the saturation can be computed algorithmically from the semigroup $\\Gamma$ and the partial gcds $d_j$.","When $\\Gamma = \\langle 2, \\gamma_2 \\rangle$ with $\\gamma_2$ odd and the characteristic condition holds, every $(\\Gamma,B)$-closed $A$ is already Lipschitz saturated, with no extra monomials needed.","The worked examples show that the final minimal generating set is often smaller than the raw gap list, since products of added generators may cover some listed degrees."],"supporting_citations":[{"why":"Defines the relative Lipschitz saturation and supplies the basic containment, idempotence, and functoriality properties used throughout.","marker":"[17]"},{"why":"Supplies the valuative criterion used in Proposition 2.5 to show that higher multiples of saturated monomials remain saturated.","marker":"[22]"},{"why":"Gives the description of the kernel of $B \\otimes_R B \\to B \\otimes_A B$ as generated by $\\Delta(g(A))$, which is the starting point of the monomial computation.","marker":"[1]"}],"fun_headline_variants":["Gaps give explicit Lipschitz saturation","Semigroup gaps reveal full Lipschitz saturation","Lipschitz saturation of monomial curves made explicit","Monomial algebroid curve saturations: formula via gaps","Explicit formula for Lipschitz saturation on monomial curves"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof relies on the ground ring being noetherian and on a property called $(\\Gamma,B)$-closedness—informally, that removing all gap-degree terms from an element of $B$ leaves an element of $A$—which is verified only for specific choices of $B$ and is not characterized in general.","fun_headline_variants_meta":{"raw":{"variants":["Gaps give explicit Lipschitz saturation","Semigroup gaps reveal full Lipschitz saturation","Lipschitz saturation of monomial curves made explicit","Monomial algebroid curve saturations: formula via gaps","Explicit formula for Lipschitz saturation on monomial curves"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000785,"raw_usage":{"total_tokens":3424,"prompt_tokens":864,"completion_tokens":2560,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":480,"completion_tokens_details":{"reasoning_tokens":2481}},"tokens_in":480,"tokens_out":2560,"duration_ms":17752,"temperature":1.0,"reasoning_tokens":2481,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T11:56:40.288334+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take $R = \\mathbb{F}_2$, $\\Gamma = \\langle 2, 3 \\rangle$, $B = \\mathbb{F}_2[t]$, and $A = \\mathbb{F}_2[t^2,t^3]$, then compute $A^*_{B,R}$ directly from the definition; here $\\operatorname{char} R$ divides $\\gamma_1$, so the theorem is not asserted, and comparing the result with the predicted $A[t^{L(r)}]$ would show whether the characteristic hypothesis is essential.","supporting_citations":[{"cited_title":"Lipman, Relative Lipschitz-saturation, Amer","cited_arxiv_id":null,"evidence_quote":"Defines the relative Lipschitz saturation and supplies the basic containment, idempotence, and functoriality properties used throughout."},{"cited_title":"Swanson and C","cited_arxiv_id":null,"evidence_quote":"Supplies the valuative criterion used in Proposition 2.5 to show that higher multiples of saturated monomials remain saturated."},{"cited_title":"Altman and S","cited_arxiv_id":null,"evidence_quote":"Gives the description of the kernel of $B \\otimes_R B \\to B \\otimes_A B$ as generated by $\\Delta(g(A))$, which is the starting point of the monomial computation."}],"review_version":1}