{"id":"b4e3ee30-d1a8-4c66-bb83-04042fe177ce","arxiv_id":"1908.05917","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For every irreducible plane curve, Yano's generating function in the characteristic sequence gives the generic b-exponents along a µ-constant deformation.","lead":"This paper proves a 1982 conjecture by T. Yano: for any irreducible plane curve singularity, the generic b-exponents (data of the Bernstein-Sato polynomial) can be computed from the combinatorial characteristic sequence of the curve. The result gives the first complete computation of all Bernstein-Sato polynomial roots for a broad family of singularities, resolving a long-open problem in singularity theory.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The proof of Theorem 17.3 asserts, without proof, that the Milnor-fiber homology is the direct sum of the images j_*H_1(X_{i,t}) over rupture divisors; the semicontinuity argument and the final assembly of Yano's conjecture both depend on this unproved splitting.","rationale":"I read the paper in good faith: it is a serious attempt at Yano's conjecture, with substantial new machinery, and the argument is coherent up to one central assembly point. The reader's weakest-assumption analysis identifies precisely the step I also find most load-bearing: the direct-sum decomposition of H_1(X_t,C) in the proof of Theorem 17.3. The phrase 'all these subspaces are direct summands' is asserted rather than proved, and the subsequent conclusion that the μ cycles form a basis is not a formal consequence of the preceding propositions. This is the point at which the local dual sections are globalized, and Proposition 8.2 requires the global dual basis to have the stated property. I do not see an internal contradiction or a reason to believe the decomposition is false; the result may well be correct. But as written, a reader cannot verify the decisive step without supplying an additional argument. Since this is a missing proof rather than a demonstrated error, the appropriate verdict remains conditional, and the reader's conditional recommendation is unchanged.","tokens_in":39,"tokens_out":22508,"duration_ms":841317,"concrete_test":"For a concrete two-Puiseux-pair branch (for example, one of the characteristic sequences (4,6,2n-3) from Remark 3.1 or the examples in [4]), compute the monodromy action on the images j_*H_1(X_{i,t}) instead of on the covering spaces X_{i,t}. Concretely, determine the characteristic polynomials of the monodromy restricted to each image from the non-resonant candidates in (17.4); test that these polynomials are pairwise coprime and that their product equals the full A'Campo characteristic polynomial (14.2). Also compute the rank of each image and verify that the total rank equals μ and that the intersection of each image with the sum of the others is zero. A cheaper necessary check is to compare ∑_{i=1}^g (c_i + N_i) with μ; if the sum exceeds μ, the unproved rank computation for the images is unavoidable.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing step occurs in the proof of Theorem 17.3: 'since we have exactly μ cycles and all these subspaces j_*H_1(X_{j,t}) are direct summands in H_1(X_t,C), one has H_1(X_t,C) = ⊕_j j_*H_1(X_{j,t}).' Having μ cycles lying in the union of the images does not imply that the sum is direct or that the images exhaust H_1. Proposition 14.2 guarantees, for each non-resonant σ_{i,ν}, an eigenvector inside j_*H_1(X_{i,t}), but it does not control overlaps between different subspaces. Moreover, Proposition 14.1 computes the Betti number of the covering space X_{i,t}, not the rank of the image of j_*; that rank can be much smaller than c_i + N_i, so the decomposition is not a formal dimension count. This equality is exactly what is needed to conclude that the constructed cycles form a global basis dual to the locally constant geometric sections, which is the hypothesis dim_C H^1_{γ_λ}(y) = 1 used in Proposition 8.2. Without the basis, the upper-semicontinuity theorem is not applicable and the final combination step in Theorem 17.5 collapses. The gap is therefore not cosmetic: it is the only place where the local sections from the individual rupture divisors are assembled into a global dual basis.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper claims a proof of Yano's conjecture for irreducible plane curve singularities: for any such germ with characteristic sequence (n, beta_1, ..., beta_g), there exists a mu-constant deformation such that the generic b-exponents are exactly the candidates in (17.4), equivalently the exponents read off from Yano's generating function R((n, beta_1, ..., beta_g), t). The proof follows Varchenko's framework: it introduces geometric and elementary sections of the cohomological Milnor fibration, reformulates b-exponents via the Brieskorn lattice and Malgrange's theorem, develops an asymptotic expansion of periods from a resolution of singularities, constructs dual locally constant geometric sections over rupture divisors, and finally assembles them through Teissier's monomial curve and semigroup-constant deformations.","tokens_in":1690,"tokens_out":1778,"duration_ms":103177,"significance":"If the proof can be completed, the result settles a 1982 conjecture of Yano and generalizes the previously known cases of one Puiseux pair and of two Puiseux pairs under distinct-eigenvalue hypotheses. A notable strength is that the candidate b-exponents are not fitted: they are derived a priori from the characteristic sequence via A'Campo's formula and Yano's generating function, and the paper gives explicit geometric constructions aimed at showing these candidates are realized. However, the manuscript's central assembly step is not fully proved, and the semicontinuity theorem and final conclusion depend on that step. The paper therefore represents a substantial and potentially correct contribution, but it is not in its current form a complete proof.","major_comments":[{"comment":"The proof asserts: since we have exactly mu cycles and all these subspaces j_*H_1(X_{j,t},C) are direct summands in H_1(X_t,C), one has H_1(X_t,C) = direct sum of j_*H_1(X_{j,t},C) over rupture divisors. This direct-sum decomposition is load-bearing and is not proved. Proposition 14.2 only produces, for each non-resonant sigma_{i,nu}, an eigenvector inside j_*H_1(X_{i,t},C) that is dual to the corresponding locally constant section within that subspace; it does not control intersections or linear relations among the images for different rupture divisors. Having mu cycles whose union spans is not enough to conclude that the sum is direct or exhaustive. Proposition 14.1 computes the Betti number of the covering space X_{i,t}, not the rank of the induced map j_*: H_1(X_{i,t},C) to H_1(X_t,C); that rank can be strictly smaller than b_1(X_{i,t}). Without a proof of the global decomposition, the hypothesis in Proposition 8.2 is not verified, and the upper-semicontinuity argument, and hence the final theorem, does not follow. This gap must be repaired by proving the global basis statement directly or by replacing the argument with one that does not rely on this unproved splitting.","section":"17, proof of Theorem 17.3"},{"comment":"The verification that A^omega_{sigma_{i,nu}-1,0} is not a section of S_{sigma_{i,nu}-2} is incomplete. The text argues from the inequality sigma_{i,nu}(omega)-1 < sigma_{i,0}(omega), but the required statement is that this particular section, whose label exponent is sigma_{i,nu}-1, cannot be expressed as a combination of sections of exponent sigma_{i,nu}-2. Since S_{sigma-2} is contained in S_{sigma-1}, an inclusion into S_{sigma-2} does not force the section itself to have exponent below sigma_{i,0}(omega)-1. A filtration argument or an explicit comparison of the locally constant geometric sections is needed. As written, the invocation of Lemma 10.6 is not justified.","section":"17, proof of Theorem 17.5"},{"comment":"The final paragraph says that upper-semicontinuity can be used to apply the argument to all candidate b-exponents simultaneously, but the deformation-theoretic mechanism is not spelled out. Proposition 17.4 produces, for each individual sigma_{i,nu}, a mu-constant deformation for which the corresponding locally constant section is generically non-zero. It is not explained how these separate deformations are merged into a single mu-constant deformation, nor why upper-semicontinuity guarantees that all candidates from (17.4) are realized for a generic fiber of one common deformation. This is the last step of the proof of Yano's conjecture and needs a precise argument.","section":"17, proof of Theorem 17.5, final combination"}],"minor_comments":[{"comment":"The sentence 'Finally, define the integers m_i := beta_i/e_i and m_i := beta_i/e_i' contains an apparent duplication; one of the two displayed definitions should be corrected.","section":"16"},{"comment":"The notation sigma_{i,nu} in (17.4) should be explicitly defined at that point, or the reader should be referred back to (10.5) with omega = dx wedge dy and to the semigroup notation introduced in Section 16.","section":"17, Eq. (17.4)"},{"comment":"The phrase 'since the pull-back of omega has exceptional support' is unclear: the intended meaning is presumably that the relevant contribution of the pulled-back form is supported near E_i, but the wording should be clarified.","section":"17, Proposition 17.4 proof"},{"comment":"The notation in Proposition 8.2 is introduced without comment and clashes with the earlier notation for monodromy eigenspaces; a brief notational reminder would help.","section":"8"},{"comment":"There are minor formatting inconsistencies in the references, including some author names and accents; these should be harmonized during revision.","section":"references"}],"recommendation":"major_revision","confidential_remarks":"The stress-test concern in the reader's report is legitimate and lands exactly on the proof of Theorem 17.3. The direct-sum splitting of H_1(X_t,C) is not a cosmetic omission; it is the hinge that connects the local duality statements from Section 14 to the global semicontinuity criterion of Section 8. Without this step, the upper-semicontinuity theorem and the final combination in Theorem 17.5 are unsupported. I would not recommend acceptance until this is proved or replaced. The paper is long and the revision should also make the simultaneous realization of all candidates in Theorem 17.5 explicit."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper is a genuine attempt at Yano's conjecture in full generality, not just the two-pair case. What's new is that it drops the pairwise-distinct eigenvalue assumption by constructing dual locally constant sections via vanishing cycles associated to each rupture divisor, then uses a semicontinuity argument in the style of Varchenko. The author knows the literature and builds on Teissier's monomial curve to produce µ-constant deformations with controlled divisorial valuations. That part is thoughtful and should be credited.\n\nThe soft spot is exactly where the reader says it is. In the proof of Theorem 17.3, the author asserts that because there are µ cycles and the subspaces j_*H_1(X_{i,t}) are direct summands, the homology is the direct sum over rupture divisors. The 'direct summand' claim is not proved, and it is load-bearing: without it, the constructed cycles don't form a global basis dual to the locally constant sections, and Proposition 8.2 does not apply. Proposition 14.2 gives duality inside each j_*H_1(X_{i,t}), not control over intersections between different subspaces. The dimension count in the text doesn't go through because Proposition 14.1 computes the Betti number of the covering, not the rank of its image in H_1(X_t). So the stress-test note is right to call this the central gap.\n\nIs it fatal? I don't know. It might be fixable by a more careful argument using the structure of the minimal resolution, and the rest of the paper is coherent enough that I would not desk-reject it. A serious referee should dig into this step. If it holds, the result is a major advance.\n\nThe paper is for specialists in singularity theory, especially people working on Bernstein-Sato polynomials and equisingular deformations. I'd bring it to a reading group, and I'd cite it if the gap gets closed. Recommendation: send it to peer review.","headline":"A serious full-generality proof of Yano's conjecture, conditional on a missing direct-sum decomposition in Theorem 17.3 that the semicontinuity argument leans on.","tokens_in":35046,"tokens_out":1693,"would_cite":true,"duration_ms":17655,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14H20","32S40","14B05","32S25"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves Yano's 1982 conjecture: generic b-exponents of irreducible plane curve singularities are computed from the semigroup by an explicit generating function.","keywords":["Bernstein-Sato polynomial","b-exponents","plane curve singularities","Yano conjecture","semigroup of values","Milnor fibration","resolution of singularities","semicontinuity"],"falsifier":"Check the asserted decomposition directly for a plane branch with at least three Puiseux pairs: compute, from the resolution dual graph and the formula in Proposition 14.1, the dimensions of the subspaces $j_*H_1(X_{i,t},\\mathbb{C})$ for each rupture divisor $E_i$, and test whether their pairwise intersections are zero and their sum is the Milnor number $\\mu$. If the dimensions do not add up, or if two subspaces share a monodromy eigenspace, Theorem 17.3's basis of eigenvectors does not exist and the proof of the conjecture fails.","tokens_in":34013,"feed_emoji":"📐","tokens_out":8816,"duration_ms":84905,"temperature":0.7,"pith_summary":"Yano's conjecture, proposed in 1982, predicts that the generic b-exponents of an irreducible plane curve singularity are determined by the curve's semigroup alone. This paper claims to prove the conjecture: for any irreducible plane branch, there is a $\\mu$-constant (topologically trivial) deformation such that for generic fibers the full set of b-exponents is given by an explicit generating function built from the resolution data. The b-exponents are normally not topological invariants and can vary in topologically trivial families, so the statement pins down their generic behavior completely. If the proof is correct, a 40-year-old conjecture is settled and the generic b-exponents of every plane branch can be read off from its characteristic sequence.","feed_headline":"Yano's conjecture proved for all irreducible plane curves","feed_subtitle":"Generic nearby curves get all their b-exponents from the semigroup's resolution data via one formula.","key_machinery":"The argument runs through four linked pieces. Malgrange's theorem identifies the b-exponents with eigenvalues of $t\\partial_t$ on the saturated Brieskorn lattice, and Varchenko's elementary sections reinterpret those eigenvalues as exponents in asymptotic expansions of period integrals. A resolution of the singularity, followed by semi-stable reduction, turns each period into a sum over exceptional divisors of terms $t^{\\sigma_{i,\\nu}-1}$ multiplied by integrals of multivalued forms $R_{i,\\nu}(\\omega)$ on the punctured projective line; a Deligne-Mostow non-vanishing criterion shows these forms give non-zero locally constant geometric sections when the exponent is non-resonant, meaning the associated numbers $\\varepsilon_{j,\\nu}(\\omega)$ are not integers. The new structural input is a duality statement: for each rupture divisor $E_i$ with $\\chi(E_i^\\circ)=-1$, the vanishing cycle lying over $E_i$ is the unique eigenvector dual to the section $A^\\omega_{\\sigma_{i,\\nu}-1,0}$, and these pairs assemble into a global basis of eigenvectors. A rupture divisor is an exceptional component whose complement has negative Euler characteristic, i.e. a branching point of the resolution dual graph, and it is the combinatorial carrier of each candidate exponent.","core_discovery":"Let $f:(\\mathbb{C}^2,0)\\to(\\mathbb{C},0)$ define an irreducible plane curve with semigroup $\\Gamma=\\langle\\beta_0,\\dots,\\beta_g\\rangle$. The central theorem (Theorem 17.5) states that there is a $\\mu$-constant deformation of $f$ such that for generic fibers the b-exponents are exactly $$\\bigcup_{i=1}^g \\left\\{\\sigma_{i,\\nu}=\\frac{m_i+n_1\\cdots n_i+\\nu}{n_i\\beta_i}\\;\\bigg|\\; 0\\le \\nu