{"id":"76047176-fd4f-4e90-9696-9889de9ca959","arxiv_id":"2608.00462","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The authors prove Dubrovin, Liu, Yang, and Zhang's conjecture: for rank-one cohomological field theories, the Dubrovin-Zhang hierarchy has a constant Poisson bracket if and only if the underlying theory is a triple Hodge class with the Calabi-Yau condition.","lead":"This paper proves a 2016 conjecture about the Poisson brackets attached to Dubrovin-Zhang hierarchies for rank-one cohomological field theories. It shows the bracket is constant exactly for the triple Hodge classes with the Calabi-Yau condition, a family tied to known integrable systems like the fractional Volterra hierarchy.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Converse hinges on Prop. 3.2: the constant bracket for triple Hodge integrals is imported from the fractional Volterra hierarchy via a non-invertible change of times, with uniqueness deferred to [LYZZ22, Prop. 3.9]; if that transfer fails, Q(e1,e3) in Thm.","rationale":"Good-faith reading: the paper's main new contribution is the converse direction, Theorem 5.1. Given formula (24) for P in terms of the Miura transformation, the strategy is sound: constancy forces all nonconstant coefficients to vanish, and the triangular system with minimal h then gives a contradiction unless the higher e_{2h-1} vanish. The algebraic parts of Sections 4-5 appear coherent: Lemma 4.4's cancellation is checkable, Lemma 5.8's identity by vanishing on hyperplanes is plausible, and Lemma 5.5's final values match the expected Bernoulli-number pattern. Credit is due for the extensive new computation in Section 5 and for the honest labeling of Prop. 3.2 as a guide rather than a full derivation. The weakest point is precisely Proposition 3.2. The transfer from FVH is non-invertible and uses formally weaker Virasoro constraints, with uniqueness cited to [LYZZ22, Prop. 3.9]. If that uniqueness theorem is exactly as needed, the logic is in fact sound: a weaker system with a unique solution still identifies the DZ tau function, because the DZ tau function satisfies the weaker system. But the paper does not reproduce the statement or the verification, and no independent computation from formula (24) is supplied for the triple Hodge specialization. Since the Q(e1,e3)=0 step in Theorem 5.1 rests entirely on this, the converse is conditional on an external theorem. This is not an internal inconsistency; it is a provenance and verification gap. The reader's weakest assumption identifies the same point, and I agree. The verdict should remain conditional: if [LYZZ22, Prop. 3.9] is accepted at face value, the proof goes through; otherwise the base point of the triangular argument is not established.","tokens_in":23267,"tokens_out":16036,"duration_ms":139883,"concrete_test":"Using the paper's own formula (24) and the B-class definition (12), compute the Poisson bracket P for the specialization e_{2k-1}=0, k>=3 and compare its w-dependent coefficients with the right-hand side of Eq. (4) through genus 3 (orders epsilon^2, epsilon^4, epsilon^6). If any nonconstant coefficient is nonzero, Prop. 3.2 is falsified and the Q(e1,e3)=0 step in Thm. 5.1 loses its base point; if the coefficients match, the FVH transfer is independently supported for these orders and the conditional verdict is strengthened.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing input is Proposition 3.2, and the paper's own proof labels it 'a short guide' to [LYZZ21, Lemma 3.9] and [LYZZ22]. The proof of Theorem 5.1 (around Eq. (69)) needs the coefficient of epsilon^{2h} w_1^{h-1} partial_x^{h+2} (and the analogous even-h coefficient) to consist of a single linear term in the minimal non-vanishing e_{2h-1} plus an 'unknown polynomial' Q(e1,e3); it then invokes Prop. 3.2 to conclude Q(e1,e3)=0, because in the triple Hodge specialization P is constant. This is exactly where the FVH transfer is used: the bracket is obtained through a non-invertible, infinite change of time variables, and the authors concede that [LYZZ21, LYZZ22] work with infinite linear combinations of flows and a formally weaker Virasoro system, relying on uniqueness in [LYZZ22, Prop. 3.9] to identify the solution. If that identification is not valid at the required level of generality, the coefficient polynomial in the triple Hodge case is not known to vanish, so the triangular system proving e_{2h-1}=0 has no verified zero base point. The paper provides no independent confirmation (e.g., a direct evaluation of (36) from formula (24)) and no machine-checked verification; the step is a citation. This is not a question of internal consistency but of provenance: the converse is only as secure as the imported uniqueness theorem.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves Conjecture 3.1 of [DLYZ16] for rank-one cohomological field theories. After recalling that every rank-one CohFT with a flat unit is written as Cl_{g,n}=exp(Σ(-1)^{i-1}(i-1)!p_i ch_i) and parametrized by e_{2h-1}, the authors use the DR/DZ equivalence to write the Poisson bracket P of the associated Dubrovin-Zhang hierarchy as P=L^{-1}∂x(L^{-1})*, where L is the linearization of the Miura map defined by B-class intersection numbers. The main theorem states that if P is constant in w_i for i≥1, then e_{2h-1}=0 for all h≥3, so Cl is a triple Hodge class Λ(r1)Λ(r2)Λ(r3) with r1r2+r1r3+r2r3=0. The proof isolates for every h≥3 a specific coefficient of P (ϵ^{2h}w1^{h-1}∂x^{h+2} for odd h, and ϵ^{2h}w1^{h-2}w2∂x^{h+1} for even h), shows via Lemma 5.5 that its leading term in e_{2h-1} is nonzero, and uses Proposition 3.2 to subtract all terms depending only on e1,e3. The converse direction, including the explicit formula (4), is imported from the fractional Volterra hierarchy literature and stated as Proposition 3.2.","tokens_in":23535,"tokens_out":15258,"duration_ms":134963,"significance":"If the main theorem and Proposition 3.2 are both fully justified, the result settles a conjecture that has been open for a decade and identifies the triple Hodge Calabi-Yau theories as the unique rank-one source of constant Poisson brackets. The paper also gives an explicit operator formula and connects the computation to Faber-type intersection numbers with explicit Bernoulli-number values. Strengths of the manuscript include the concrete coefficient computations in Lemma 5.5, the cross-check against the Λ(-r) example, and the transparent reduction of the constancy condition to a triangular system. The main concern is that the reverse implication rests on Proposition 3.2, whose proof in this paper is a guide to prior work rather than a complete verification.","major_comments":[{"comment":"Proposition 3.2 is load-bearing for the proof of Theorem 5.1: Eq. (69) uses it to conclude Q(e1,e3)=0, i.e. that the coefficient of ϵ^{2h}w1^{h-1}∂x^{h+2} vanishes in the triple Hodge Calabi-Yau specialization. The proposition is not proved in this paper; its proof is explicitly 'a short guide' to [LYZZ21, Lemma 3.9] and [LYZZ22, Prop. 3.9], and the transfer from the fractional Volterra hierarchy is via a non-invertible change of time variables and formally weaker Virasoro constraints. A referee cannot verify that the unique solution identified in [LYZZ22, Prop. 3.9] is the Dubrovin-Zhang hierarchy of the triple Hodge CohFT in the sense of Section 4. I request a self-contained proof of Proposition 3.2, or a precise statement of a theorem in the cited papers that applies directly to the DZ hierarchy together with a verification of its hypotheses. Without this, the base point of the triangular system in Theorem 5.1 is not established.","section":"§3, Proposition 3.2 and §5.3, Eq. (69)"}],"minor_comments":[{"comment":"In the arXiv text the formula is typeset as a product of S-factors times ∂x; the intended expression is a quotient with ∂x in the numerator and the three S-factors in the denominator. Please correct the typesetting.","section":"Eq. (4) and abstract"},{"comment":"The rescaling ϵ̃=√(r1r2r3)ϵ is used, but when e3=r1r2r3 vanishes this rescaling degenerates. Remark 3.3 treats the limit, but a sentence stating that the degenerate cases follow by continuity or by the direct computation in Section 4.4 would make the proposition unambiguous.","section":"§3, proof of Proposition 3.2"},{"comment":"The formula π_*∏_{i=1}^{n+1}(1-b_iψ_i)^{-1}=Σ_{i=1}^{n+1}b_i∏(1-b_iψ_i)^{-1} is stated without justification; adding a one-line derivation or a reference would improve readability.","section":"§4.2, Lemma 4.4"},{"comment":"The simplification of the two displayed integrals to -B_{2h}h!(h-2)/(3·2^{h-1}·(2h)!) is not shown; including the intermediate algebra would make the even-h nonvanishing easier to verify.","section":"§5.2.2, Eq. (45)"}],"recommendation":"major_revision","confidential_remarks":"The core computational argument is explicit and the nonvanishing Lemma 5.5 appears sound. My reservation is entirely about the provenance of Proposition 3.2, which is the base point of the converse. If the authors can supply a complete proof of that proposition in the revision, or clearly locate an existing theorem with the exact hypotheses needed for the DZ bracket, I would view the paper as suitable for acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is the converse half of the DLYZ16 conjecture, and it is a real result. The forward direction (triple Hodge Calabi-Yau gives constant bracket) was already in LYZZ21/LYZZ22; what's new is Theorem 5.1, which shows constancy forces e_{2h-1}=0 for h>=3. The proof is a serious computation: write the DZ bracket via the DR/DZ Miura transformation, expand in epsilon, and show the first non-vanishing higher e gives a nonzero coefficient. The use of Proposition 4.8 and the L operator is clean, and Lemma 5.5 reduces the key coefficients to Faber socle numbers with explicit Bernoulli factors. I did not find a gap in the coefficient extraction; the triangular structure works.\n\nThe soft spot is Proposition 3.2. The constant bracket for the triple Hodge case is imported from the fractional Volterra hierarchy via a non-invertible change of time variables, and the proof is explicitly a short guide to LYZZ21/LYZZ22. The step Q(e1,e3)=0 in Theorem 5.1 depends on that import being the actual DZ bracket on the nose. If the uniqueness in LYZZ22, Prop. 3.9 does not cover the needed generality, the base point of the induction is not verified. The authors know this and give reasons to trust it, but they do not provide an independent check, such as a direct evaluation of Eq. (36) or a machine-checked computation. This is a provenance issue, not an internal inconsistency. I would want a referee to look carefully at Prop. 3.2 and the cited uniqueness theorem.\n\nOverall, the paper deserves serious refereeing. It resolves a conjecture, the new part is definitely new, and the main computation is explicit and cross-checked on Lambda(-r). The soft spot is real but narrow and clearly flagged; it is a fixable or verifiable gap, not a conceptual one. I would send it to a competent referee rather than desk reject.","headline":"The converse of the DLYZ conjecture is proven by explicit DR/DZ computations; the only real question is whether the imported FVH transfer in Prop. 3.2 is solid enough to carry the base point.","tokens_in":24127,"tokens_out":2146,"would_cite":true,"duration_ms":19132,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14H10","14N35","37K10"],"pacs":[],"model":"deepseek-v4-flash","headline":"The Poisson bracket of a rank-one Dubrovin-Zhang hierarchy is constant exactly for the Calabi-Yau triple Hodge class $\\Lambda(r_1)\\Lambda(r_2)\\Lambda(r_3)$, with an explicit bracket formula.","keywords":["Dubrovin-Zhang hierarchies","cohomological field theories","Hodge integrals","Poisson brackets","triple Hodge classes","Calabi-Yau condition","fractional Volterra hierarchy","moduli of curves"],"falsifier":"For a rank-one theory with $e_5\\neq0$ and all $e_7,e_9,\\ldots$ zero, compute the coefficient of $\\epsilon^6 w_1^2\\partial_x^5$ in its Dubrovin-Zhang bracket; Example 5.2 reduces it to $120e_5\\int_{M_{3,3}} ch_5\\,\\mathrm{Coeff}[a_1a_2a_3^2]B^0_{3,3}$, which Lemma 5.5 says is nonzero. A theory in which that coefficient vanishes while $e_5\\neq0$ would refute Theorem 5.1.","tokens_in":23015,"feed_emoji":"🧮","tokens_out":12972,"duration_ms":100656,"temperature":0.7,"pith_summary":"This paper proves a conjecture about when the Dubrovin-Zhang integrable hierarchy attached to a rank-one cohomological field theory has a Poisson bracket with constant coefficients. The answer is that this happens exactly for the triple Hodge classes $\\Lambda(r_1)\\Lambda(r_2)\\Lambda(r_3)$ whose parameters satisfy the Calabi-Yau condition $r_1r_2+r_1r_3+r_2r_3=0$, and in that case the bracket has an explicit closed form. This matters because constant brackets are structurally special: they are the cases tied to the fractional Volterra hierarchy and to KP integrability, and the explicit formula makes the whole Poisson structure computable. The proof writes the bracket through a Miura transformation from the double-ramification hierarchy and shows that the first non-constant coefficients force every higher odd Hodge parameter to vanish.","feed_headline":"Only Calabi-Yau triple Hodge theories have constant Poisson brackets","feed_subtitle":"A constant bracket pins the theory to the Calabi-Yau triple Hodge family, with an explicit closed bracket.","key_machinery":"The load-bearing mechanism is the DR/DZ equivalence. For a rank-one theory the Dubrovin-Zhang hierarchy is obtained from the double-ramification hierarchy, whose Poisson bracket is simply $\\partial_x$, by a Miura transformation $L$, so the bracket is $P=L^{-1}\\partial_x(L^{-1})^*$ (Eq. (24)). The transformation is expressed through intersection numbers of the CohFT class with tautological classes $B^m_{g,n}$ on moduli spaces of stable rooted trees. The proof then reads off the top-derivative coefficients of $P$ in normal form: for odd $h$, the coefficient of $\\epsilon^{2h}w_1^{h-1}\\partial_x^{h+2}$; for even $h$, the coefficient of $\\epsilon^{2h}w_1^{h-2}w_2\\partial_x^{h+1}$. Lemma 5.5 evaluates these as nonzero multiples of Bernoulli numbers, producing a triangular system that forces each $e_{2h-1}$, $h\\ge3$, to vanish.","core_discovery":"The central discovery is Theorem 5.1: if the Poisson bracket $P$ of the Dubrovin-Zhang hierarchy associated with a rank-one cohomological field theory is constant in the dependent variables $w_i$, $i\\ge1$, then $e_{2h-1}=0$ for all $h\\ge3$. Equivalently, the theory is $C_{g,n}=\\Lambda(r_1)\\Lambda(r_2)\\Lambda(r_3)$ with $r_1r_2+r_1r_3+r_2r_3=0$. Combined with Proposition 3.2, which exhibits the explicit constant bracket (4) for exactly this family, this proves Conjecture 3.1 of [DLYZ16] in full. The constancy condition leaves only two free parameters, $e_1$ and $e_3$.","pith_inferences":["A testable extension is to ask whether requiring the bracket to be constant only up to a fixed differential order still forces the same vanishing; the $e_5$ obstruction suggests the effect already appears at the first non-vanishing odd Chern character.","The Bernoulli-number coefficients supply a quantitative obstruction spectrum, so one could measure how far a given hierarchy is from constant-bracket form by the size of its first surviving coefficient.","The same B-class intersection machinery may apply to partial cohomological field theories or to constancy of higher Poisson brackets, where the triangular structure would recur.","If the result is read together with the known KP integrability of the Calabi-Yau triple Hodge family, it suggests that within rank-one theories 'constant Poisson bracket' and 'KP-integrable Dubrovin-Zhang hierarchy' coincide."],"forward_implications":["Every rank-one theory outside the family $\\Lambda(r_1)\\Lambda(r_2)\\Lambda(r_3)$ with $r_1r_2+r_1r_3+r_2r_3=0$ has a non-constant Dubrovin-Zhang bracket, with the first obstruction given by a computable Bernoulli-number coefficient.","For the surviving family the bracket is $P=\\partial_x/[S(\\sqrt{r_1r_2/r_3}\\,\\epsilon\\partial_x)S(\\sqrt{r_2r_3/r_1}\\,\\epsilon\\partial_x)S(\\sqrt{r_3r_1/r_2}\\,\\epsilon\\partial_x)]$, so the full Poisson structure is explicit.","The limiting cases recover known constant brackets: $\\Lambda(-r)$ gives $\\partial_x/S(\\sqrt r\\,\\epsilon\\partial_x)^2$, and $r\\to0$ gives the KdV bracket $\\partial_x$.","Constancy of the Poisson bracket is therefore a sharp criterion that singles out the Calabi-Yau triple Hodge class among all rank-one cohomological field theories."],"supporting_citations":[{"why":"Poses the conjecture and constructs the Dubrovin-Zhang hierarchy for Hodge-type theories; it is the statement being proved.","marker":"[DLYZ16]"},{"why":"Supplies the Hodge-to-fractional-Volterra correspondence used in Proposition 3.2 to obtain the explicit constant bracket (4).","marker":"[LYZZ21]"},{"why":"Proves the loop equation and the uniqueness result that lets the paper identify the quasi-Miura transformation with the Dubrovin-Zhang one.","marker":"[LYZZ22]"},{"why":"Provides the DR/DZ equivalence relations underlying the Miura-transformation formula used in Section 4.","marker":"[BS24]"},{"why":"Gives the explicit Miura transformation and the conjugation formula $P=L^{-1}\\partial_x(L^{-1})^*$ used throughout.","marker":"[BLS26]"},{"why":"Classifies rank-one cohomological field theories, giving the Hodge-exponential form (1) from which all computations start.","marker":"[Tel12]"},{"why":"Supplies the relations among Chern characters of the Hodge bundle used to reduce the intersection numbers in the obstructions.","marker":"[Mum83]"},{"why":"Establishes the socle intersection-number formula quoted as Lemma 5.11, which yields the nonzero Bernoulli-number coefficients.","marker":"[GP98]"}],"fun_headline_variants":["Constant Poisson brackets force Calabi-Yau triple Hodge classes","Calabi-Yau triple Hodge is the only source of constant brackets","Constant bracket ⇒ Calabi-Yau triple Hodge theory","Calabi-Yau condition is the key to constant Poisson brackets","Constant bracket pinpoints Calabi-Yau triple Hodge"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof that the triple Hodge Calabi-Yau hierarchy really has the constant bracket (4) is inherited from earlier work through a quasi-Miura transformation built from a tau function, and that earlier work uses infinite linear combinations of flows and a formally weaker system of Virasoro equations; if that identification is not exactly the Dubrovin-Zhang hierarchy, the step $Q(e_1,e_3)=0$ in the proof of Theorem 5.1 has no basis.","fun_headline_variants_meta":{"raw":{"variants":["Constant Poisson brackets force Calabi-Yau triple Hodge classes","Calabi-Yau triple Hodge is the only source of constant brackets","Constant bracket ⇒ Calabi-Yau triple Hodge theory","Calabi-Yau condition is the key to constant Poisson brackets","Constant bracket pinpoints Calabi-Yau triple Hodge"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001083,"raw_usage":{"total_tokens":4437,"prompt_tokens":764,"completion_tokens":3673,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":380,"completion_tokens_details":{"reasoning_tokens":3587}},"tokens_in":380,"tokens_out":3673,"duration_ms":21594,"temperature":1.0,"reasoning_tokens":3587,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T15:19:04.055776+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"For a rank-one theory with $e_5\\neq0$ and all $e_7,e_9,\\ldots$ zero, compute the coefficient of $\\epsilon^6 w_1^2\\partial_x^5$ in its Dubrovin-Zhang bracket; Example 5.2 reduces it to $120e_5\\int_{M_{3,3}} ch_5\\,\\mathrm{Coeff}[a_1a_2a_3^2]B^0_{3,3}$, which Lemma 5.5 says is nonzero. A theory in which that coefficient vanishes while $e_5\\neq0$ would refute Theorem 5.1.","supporting_citations":[],"review_version":2}