{"id":"ce439b25-3dc2-4b99-8a75-f897dc571a34","arxiv_id":"2506.02540","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The reduced DR hierarchy built from spin-weighted strata of residueless meromorphic differentials with two zeros equals the BKP hierarchy up to an explicit coordinate transformation.","lead":"Spin refinements of moduli spaces of residueless meromorphic differentials give a system of PDEs that matches the BKP hierarchy, a classical reduction of the KP equation, after an explicit rescaling. The result extends the known KP correspondence to the spin setting and supports the pattern that spin versions of geometric problems are governed by BKP-type integrable hierarchies.","discovery_kind":"unification","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The BKP identification rests on Lemma 3.5, whose allegedly unconditional proof imports Wong's assumption-based splitting formulas; if those fail, the initial datum Qspin_{2,2} changes and Theorem 5.1 is not established.","rationale":"The reader's weakest-assumption diagnosis matches my reading of the paper. The architecture of the argument is otherwise sound: the coordinate transformation in Theorem 5.1 is consistent with properties (52)-(57) and (59)-(64), and the reconstruction theorem is an independent Lax-formalism statement. The single load-bearing geometric input is the genus-2 computation of Lemma 3.5. The first proof is admittedly conjectural, and the second proof, though longer and more explicit, still imports Wong's splitting formulas, which the paper itself flags as assumption-based. Since the rest of the reconstruction depends only on Qspin_{2,2}, a failure or unproven status of Lemma 3.5 blocks the unconditional identification with BKP. I therefore see no reason to move the reader's CONDITIONAL verdict; the concern is real but not a demonstrated contradiction.","tokens_in":55660,"tokens_out":12414,"duration_ms":124762,"concrete_test":"Re-derive Proposition 3.7 for the specific twisted graphs contributing to Lemma 3.13 and Proposition 3.15 directly from CSS21 Lemma 5.7 and the projection formula, without invoking Wong's Propositions 5.9/5.10; then recompute the H-integrals in Proposition 3.15 and the final values 37/1152 and 7/5760 of Lemma 3.5. If this direct derivation cannot be completed without an unproved assumption, or if the recomputed values differ, the conditional status of Theorem 5.1 is confirmed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Theorem 5.1 follows from Theorem 5.3, whose only geometric input from Section 3 is the initial datum Qspin_{2,2}, fixed by Lemma 3.5. The first proof of Lemma 3.5 explicitly uses the CSS21 spin DR formula (21) for g=2, which the paper marks as unproven (Section 3.2.1). The second proof is intended to avoid (21), but it is not unconditional: Proposition 3.7 is proved from Wong's Propositions 5.9/5.10 and Corollary 1.3, and Proposition 3.8 is obtained by the same 'taking differences' procedure, while Section 1.2 states that Wong's algorithm is 'based on a few assumptions'. These propositions are then used in the systems for A,B,C in Lemma 3.13, in Proposition 3.14, and throughout Proposition 3.15. If any of Wong's assumptions fails for the two- or three-entry level graphs used there, the values 37/1152 and 7/5760 in Lemma 3.5 change; since Qspin_{2,2} is the unique nontrivial datum in the reconstruction theorem, the claimed equality with BKP would not be established. The paper does not isolate which of Wong's assumptions are needed for these specific graphs, nor does it provide an independent verification of the two intersection numbers.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper constructs a partial cohomological field theory from the spin-parity-weighted classes [H^res_g(2α_1,...,2α_n)]_spin of residueless meromorphic differentials with even vanishing orders, applies the double-ramification (DR) hierarchy construction, and reduces to the locus of differentials with two zeros. The main theorem (Theorem 5.1) states that, after a change of variables v_α = -w_{2α-1}/(2α-1), t_β = (2β-1)T_{2β-1}, and ε² = 2, the reduced DR hierarchy coincides with the BKP hierarchy. The proof combines a new reconstruction theorem for BKP from limited initial data (Theorem 5.3) with explicit computations of the first nontrivial differential polynomials in genus 0, 1, and 2. The genus-2 computation (Lemma 3.5) is the key geometric input.","tokens_in":55916,"tokens_out":3505,"duration_ms":36592,"significance":"If the main theorem is established, it gives a sharp spin refinement of the Buryak–Rossi–Zvonkine result relating residueless differential strata to KP, and it fits the emerging pattern that spin versions of KP-governed problems are governed by BKP. The BKP hierarchy is defined independently via the Lax constraint (Section 4.2), not by fitting to moduli geometry, so there is no circularity in the overall strategy. The reconstruction theorem (Theorem 5.3) is a genuine uniqueness result and is stronger in terms of required input than the analogous KP reconstruction in [BRZ24]. The paper is also transparent about where it relies on external conjectural or assumption-based input, and the explicit intersection-theoretic computations, several checked with admcycles, are a useful contribution. However, the genus-2 initial datum that feeds the reconstruction rests on input whose unconditional status is not fully established, and this blocks acceptance in the present form.","major_comments":[{"comment":"The first proof of Lemma 3.5 explicitly uses the genus-2 case of the conjectural spin DR formula (21), which the paper itself identifies as unproven (Section 3.2.1, citing CSS21 Assumption 1.3). Since Lemma 3.5 determines the potential Q^spin_{2,2}, and Q^spin_{2,2} is the unique nontrivial initial datum in the reconstruction theorem, this proof cannot serve as an unconditional verification of the two intersection numbers 37/1152 and 7/5760. The first proof should either be explicitly labelled as conditional or removed from the proof of Proposition 2.5 and Theorem 5.1.","section":"§3.2.1, Eq. (21) and Lemma 3.5"},{"comment":"The second proof of Lemma 3.5 is intended to avoid the conjectural formula (21), but it imports Propositions 3.7 and 3.8 from Wong's algorithm [Won24], and Section 1.2 states that this algorithm is 'based on a few assumptions'. The paper does not isolate which of Wong's assumptions are needed for the specific two- and three-entry level graphs used in Lemma 3.13, Proposition 3.14, and Proposition 3.15, nor does it provide an independent verification of the two intersection numbers. If any of those assumptions fails, the values 37/1152 and 7/5760 change and Theorem 5.1 is not established. The unconditional status of these computations needs to be repaired, for example by proving the needed special cases of Wong's algorithm or supplying an independent genus-2 computation.","section":"§3.2.2, Propositions 3.7 and 3.8"},{"comment":"The dependence of the main theorem on Lemma 3.5 is load-bearing: Theorem 5.1 follows from Theorem 5.3 together with properties (59)–(64), and the only nontrivial geometric input among those properties is the value of Q^spin_{2,2} fixed by Lemma 3.5. Because both available proofs of Lemma 3.5 rely on conjectural or assumption-based input, the main theorem is currently conditional on external results that are not proved in the paper. The authors should either make the theorem explicitly conditional, or supply a complete proof of the needed genus-2 spin intersection numbers.","section":"§5.1, Theorem 5.1 and Theorem 5.3"}],"minor_comments":[{"comment":"The sentence 'The stack B^res_g(α_1,...,α_n) is a moduli stack for for families of equivalence classes...' contains a duplicated 'for'.","section":"§1.1, Proposition 1.4"},{"comment":"The final sentence of the proof says 'The last equality is due to Lemma 3.10', but the displayed expression also uses equation (26); please make the reference precise.","section":"§3.2.2, Lemma 3.13"},{"comment":"The sentence 'Note that the proof of this determination uses a non-constructive argument only at the end' is vague; it would help the reader to indicate explicitly that the non-constructive step is the contradiction argument with λ in the proof of Theorem 5.3.","section":"§5.2, after Lemma 5.4"},{"comment":"The notation Tf_u in equation (65) is introduced only inside the remark; defining it before the displayed equation would make the remark easier to follow.","section":"§5.1, Remark 5.2"}],"recommendation":"major_revision","confidential_remarks":"The central claim is plausible and the overall architecture is coherent, but the paper currently presents Lemma 3.5 as proved while both proofs rely on unproven or assumption-based external results. This is a correctness-risk that should be resolved publicly before the paper is accepted, not merely acknowledged in a footnote. The reconstruction theorem and the explicit computations are valuable independently, so a careful revision that either supplies the missing unconditional verification or states the main theorem as conditional would make the contribution suitable for the journal."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The new thing here is real: the authors prove that the DR hierarchy attached to their spin-refined partial CohFT of residueless meromorphic differentials with two zeros coincides with the BKP hierarchy after an explicit rescaling, and they also prove a reconstruction result that needs less Cauchy data than the analogous KP statement in BRZ24. The paper is honest and well structured; the partial CohFT setup, the reduction to two zeros, and the Lax-based definition of BKP are all clearly written, and the reconstruction proof is long but mostly self-contained. I walked through the coordinate transformation and the structure of the reconstruction argument and found no circularity: BKP is defined from Lax operators, the geometric initial data are intersection numbers, and the theorem is a genuine identification, not a restatement.\n\nThe soft spot is exactly where the reader put it, Lemma 3.5. The two numbers 37/1152 and 7/5760 determine Q^spin_{2,2}, which is the unique nontrivial input to the reconstruction theorem. The first proof of that lemma uses the conjectural spin DR formula (21) for g=2, which the paper itself marks as unproven. The second proof is meant to be unconditional, but it pulls in Propositions 3.7 and 3.8, which are derived from Wong's algorithm, and the paper's introduction says that algorithm is based on a few assumptions. The stress-test note is right that the authors do not isolate which of Wong's assumptions are actually needed for the specific two- and three-entry level graphs they use, and they do not provide an independent verification of the two integrals. If those numbers shift, the claimed equality with BKP is not established. That is a narrow but load-bearing gap, and it is the only substantial concern I have.\n\nWho gets value from this: anyone working on DR hierarchies, spin strata, or the spin-KP/BKP correspondence will want to read it. It is a solid within-subfield advance, not a revolution, and the presentation is workmanlike rather than flashy. I would be comfortable sending it to a serious referee, with the explicit request to check Lemma 3.5, ideally by an independent computation of the two genus-2 intersections or by a careful audit of the Wong-based splitting steps. My own verdict would be conditional: the framework and reconstruction are sound, and the two intersection numbers are likely correct, but they are not yet fully secured.","headline":"A genuinely new spin-KP/BKP bridge, but the identification rests on two genus-2 intersection numbers that are not yet independently verified.","tokens_in":56463,"tokens_out":1132,"would_cite":true,"duration_ms":14717,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14H10","14H15","37K10"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that the DR hierarchy of spin-refined strata of residueless meromorphic differentials with two zeros coincides with the BKP hierarchy after an explicit rescaling.","keywords":["spin structures","meromorphic differentials","BKP hierarchy","double ramification hierarchy","cohomological field theory","moduli of curves","theta characteristics"],"falsifier":"Recompute the two integrals in Lemma 3.5 by a method that uses neither the conjectural spin DR formula (21) for $g=2$ nor Wong's algorithm; if the values $37/1152$ and $7/5760$ are not reproduced, Theorem 5.1 fails for the stated transformation.","tokens_in":55408,"feed_emoji":"🌀","tokens_out":6014,"duration_ms":54499,"temperature":0.7,"pith_summary":"The paper proves that a spin-refined version of the moduli problem governing residueless meromorphic differentials is governed by the BKP hierarchy, not merely by KP. Curves carrying such a differential with even zero and pole orders inherit a spin structure, and the paper forms the cohomology class obtained by subtracting the odd-parity locus from the even-parity locus. These classes satisfy the axioms of a partial cohomological field theory of infinite rank. Applying the double ramification (DR) hierarchy construction and restricting to differentials with exactly two zeros produces a system of PDEs that, after an explicit change of variables and the substitution $\\varepsilon^2=2$, coincides with the BKP hierarchy. The proof is built on a new reconstruction theorem showing that BKP is uniquely determined by its linear dispersionless term, its first nontrivial potential, and the commutativity and homogeneity properties of its flows.","feed_headline":"Spin-refined strata of two-zero differentials equal BKP","feed_subtitle":"Even-minus-odd spin classes form a partial CohFT whose DR hierarchy is BKP after an explicit rescaling.","key_machinery":"The load-bearing objects are the spin-parity cohomology classes $[H^{\\mathrm{res}}_g(2\\alpha_1,\\ldots,2\\alpha_n)]_{\\mathrm{spin}} = [\\text{even}] - [\\text{odd}]$, which form a homogeneous partial CohFT of infinite rank with unit $e_0$ and metric $\\eta_{\\alpha\\beta} = \\delta_{\\alpha+\\beta,-1}$. The DR hierarchy construction turns this partial CohFT into commuting Hamiltonian flows whose densities are defined by integrals of DR cycles, $\\lambda$-classes, and $\\psi$-classes against these spin strata. The decisive mechanism is Theorem 5.3, a reconstruction principle: the compatibility of the flows plus homogeneity, tau-symmetry, and translation invariance determine every polynomial $Q^{\\mathrm{spin}}_{\\alpha\\beta}$ once the linear term and the first nontrivial potential $Q^{\\mathrm{spin}}_{2,2}$ are known. This reduces the geometric verification to computing intersection numbers in genera 0, 1, and 2.","core_discovery":"The central discovery is Theorem 5.1: for the spin-refined strata, the reduced DR hierarchy $\\partial v_\\alpha/\\partial t_\\beta = \\partial_x Q^{\\mathrm{spin}}_{\\alpha\\beta}$ and the BKP hierarchy in normal coordinates $\\partial w_{2\\alpha-1}/\\partial T_{2\\beta-1} = \\partial_x R^{\\mathrm{BKP}}_{\\alpha\\beta}$ are the same system. The identification is made by $v_\\alpha = -w_{2\\alpha-1}/(2\\alpha-1)$, $t_\\beta = (2\\beta-1)T_{2\\beta-1}$, together with $\\varepsilon^2=2$. Under this substitution, the geometric data, including the genus-2 potential $Q^{\\mathrm{spin}}_{2,2}$, are sent exactly to the BKP data, including $R^{\\mathrm{BKP}}_{2,2} = \\tfrac{9}{5}w_5 - w_3^{(2)} - 3w_1w_3 + \\tfrac{1}{5}w_1^{(4)} + 3w_1w_1^{(2)} + 3w_1^3$. The argument does not require knowing the full BKP hierarchy: Theorem 5.3 reconstructs every $Q^{\\mathrm{spin}}_{\\alpha\\beta}$ recursively from $Q_{\\gamma,2}$ using only commutativity of flows and the listed structural properties, which is a genuinely new reconstruction result for BKP.","pith_inferences":["One could test the reconstruction's minimality by checking whether replacing $Q^{\\mathrm{spin}}_{2,2}$ with any other admissible potential still yields a commuting hierarchy; the proof suggests it would not, implying the genus-2 numbers are uniquely forced.","The $\\varepsilon^2=2$ substitution suggests that in any further spin/KP correspondence the spin parameter enters only through this normalization, which might serve as a check for other spin-refined hierarchies.","A fully independent computation of the two genus-2 integrals in Lemma 3.5, once the spin DR conjecture is proved, would either close the remaining gap or, if it disagrees, identify exactly where the BKP identification would break."],"forward_implications":["If Theorem 5.1 is correct, every flow of the BKP hierarchy in normal coordinates is realized geometrically as a DR flow on spin-refined strata of two-zero residueless differentials.","The BKP hierarchy is completely determined by the linear term of its dispersionless limit together with the first nontrivial potential, so other geometric constructions sharing these initial data must produce the same hierarchy.","The result gives a concrete instance of the principle that spin refinements of KP-governed enumerative problems are governed by BKP, matching the known spin Hurwitz number phenomenon.","The intersection numbers of Lemma 3.5 become fixed geometric inputs that constrain all higher coefficients of the hierarchy, so any future computation of spin stratum classes must be consistent with the BKP values."],"supporting_citations":[{"why":"Supplies the non-spin analogue: residueless meromorphic differential strata produce the KP hierarchy, and provides the overall strategy of reconstructing the hierarchy from limited data.","marker":"[BRZ24]"},{"why":"Introduces the DR hierarchy construction that assigns a compatible Hamiltonian PDE system to a CohFT or partial CohFT.","marker":"[Bur15]"},{"why":"Provides the conjectural spin DR formula (Assumption 1.3) and formula (21) used in the first proof of the genus-2 intersection numbers in Lemma 3.5.","marker":"[CSS21]"},{"why":"Supplies an algorithm for computing spin-parity fundamental classes of strata of differentials, used in the second method for Lemma 3.5 and in verification.","marker":"[Won24]"},{"why":"Provides the formula for $\\mathrm{DR}_g(a,-a)\\lambda_g$ used to reduce the genus-2 integrals in Lemma 3.5.","marker":"[BHS22]"},{"why":"Gives Hain's formula for double ramification cycles restricted to compact type, used throughout the intersection-theoretic computations.","marker":"[Hai13]"},{"why":"Defines the BKP hierarchy as a reduction of the KP hierarchy, the target object identified in Theorem 5.1.","marker":"[DJKM82]"},{"why":"Proves that the BKP constraint is invariant under odd flows of the KP hierarchy, a property needed to write BKP in odd variables.","marker":"[Zab21]"}],"fun_headline_variants":["Spin-refined strata of two-zero differentials match BKP hierarchy","Spin refinement keeps BKP hierarchy for differential strata","DR hierarchy from spin strata equals BKP after rescaling","Spin-weighted strata reduce DR hierarchy to BKP exactly","Two-zero differential spin strata yield BKP hierarchy"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The identification with BKP rests on the two genus-2 intersection numbers of Lemma 3.5 being correct: the first proof of them uses a conjectural spin double-ramification formula that the paper marks as unproven, and the second proof imports an algorithm the paper describes as based on assumptions.","fun_headline_variants_meta":{"raw":{"variants":["Spin-refined strata of two-zero differentials match BKP hierarchy","Spin refinement keeps BKP hierarchy for differential strata","DR hierarchy from spin strata equals BKP after rescaling","Spin-weighted strata reduce DR hierarchy to BKP exactly","Two-zero differential spin strata yield BKP hierarchy"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000767,"raw_usage":{"total_tokens":3416,"prompt_tokens":975,"completion_tokens":2441,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":591,"completion_tokens_details":{"reasoning_tokens":2377}},"tokens_in":591,"tokens_out":2441,"duration_ms":17764,"temperature":1.0,"reasoning_tokens":2377,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T11:22:18.535846+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the two integrals in Lemma 3.5 by a method that uses neither the conjectural spin DR formula (21) for $g=2$ nor Wong's algorithm; if the values $37/1152$ and $7/5760$ are not reproduced, Theorem 5.1 fails for the stated transformation.","supporting_citations":[],"review_version":1}