{"id":"bdee9986-cedd-48d6-9d39-9e5fdf39ef00","arxiv_id":"2607.16607","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Finite-group reduction preserves bihamiltonian and tau structures of DR/DZ hierarchies, and two inequivalent reductions of the P¹_{2,2,2,2} CohFT give two higher-genus completions of M_{1,1}, one equivalent to genus-1 topological recursion.","lead":"This paper proves that restricting an integrable hierarchy of DR/DZ type to the flows invariant under a finite group action keeps its two compatible Hamiltonian structures and its tau structure intact. It then exhibits two genuinely different reduced hierarchies on the same genus-zero geometry (the Hurwitz space M_{1,1}) — one matching genus-1 topological recursion, the other a BCFG-type modified Virasoro case.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The application's central-invariant computation and its uncited Virasoro uniqueness step are the most load-bearing unverified points; Theorem 1.1's reduction proof is more robust than the reader's flagged transfer assumption suggests.","rationale":"The reader's CONDITIONAL verdict is reasonable: the paper's structural theorem is supported by a coherent reduction argument with heavy lifting from cited preprints, while the application contains unshown computations and an uncited uniqueness statement. I did not find a demonstrated error in Theorem 1.1; the reduction proof (Lemma 4.2, Theorems 4.3, 4.11, 4.14) provides a self-contained route that bypasses the §3.4 numerical-partial transfer issue highlighted by the reader. The more load-bearing gap is in §5.2.2: the central-invariant value {1/12,1/12,1/12} is asserted without displaying the necessary computation, and the final identification with topological recursion rests on a Virasoro rigidity statement that is neither proved nor referenced. These are concrete, checkable gaps rather than demonstrated failures, so the appropriate verdict remains CONDITIONAL, not ACCEPT or REJECT. My read therefore does not change the reader's verdict.","tokens_in":30206,"tokens_out":35439,"duration_ms":370707,"concrete_test":"Compute the Z2×Z2-reduced pencil (P1^{Γ},P2^{Γ}) to order ε^3 in the DZ normal coordinates using the restriction formulas of §4.1 and the prepotential (5.7), then evaluate the central invariants using the formula in §5.1. Verify they equal {1/12,1/12,1/12} before the ε-rescaling and {1/24,1/24,1/24} after. Independently, supply a proof or precise reference for the Virasoro uniqueness theorem invoked in §5.2.2, confirming that it covers numerically partial CohFTs lacking (C4). If the uniqueness theorem does not apply, the equivalence to genus-one topological recursion is unsupported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The advertised consequence, Theorem 1.2(2) — equivalence of the rescaled Z2×Z2-reduced hierarchy to genus-one topological recursion — rests on two unverified inputs. First, §5.2.2 asserts 'One can verify similarly that the central invariants in this case are given by {1/12,1/12,1/12}' without displaying the P1,P2 expansions, the canonical coordinates, or the effect of the two rescalings (2^{1-g}Λ and ε→ε/2) on the pencil. Second, the identification with topological recursion uses an uncited uniqueness claim: 'The Virasoro constraints force them to coincide up to an automorphism...' This is a nontrivial rigidity theorem for tau functions with fixed dispersionless limit and standard Virasoro symmetries; it is not proved or referenced, and it is unclear whether it applies to numerically partial CohFTs lacking (C4). If either the central invariants are not {1/12,1/12,1/12} (before rescaling) or the Virasoro uniqueness fails in this class, the claimed equivalence to topological recursion collapses. By contrast, the main structural Theorem 1.1 is less exposed: the reduction arguments in Lemmas 4.2, Theorems 4.3, 4.11, and 4.14 work directly with the ambient hierarchy and the restriction map ι*, so they do not actually depend on the questionable §3.4 transfer assertions about general numerically partial CohFTs. The reader's weakest_assumption is therefore not the most load-bearing point.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a Γ-reduction formalism for the Dubrovin–Zhang and Double Ramification hierarchies associated to a semisimple CohFT with a numerical finite group symmetry. Section 2 introduces Γ-linearized coordinates and proves algebraic lemmas for passing to fixed loci of differential polynomials. Section 3 defines numerically partial CohFTs and associates a DR hierarchy to them, asserting commutativity and bihamiltonicity in this weakened setting. Section 4 proves the main structural theorem: for a semisimple CohFT with numerical symmetry Γ, the Γ-invariant flows of the ambient DR and DZ hierarchies form bihamiltonian hierarchies with inherited tau structures, connected by a Miura-type transformation (Theorems 4.3, 4.11, 4.14). Section 5 applies the result to the orbifold Gromov–Witten theory of P^1_{2,2,2,2}: the natural S_4 reduction has central invariants {1/6, 1/24, 1/24}, while an auxiliary Z_2×Z_2 action yields {1/12, 1/12, 1/12}; the paper then claims that the rescaled reduced hierarchy is equivalent to genus-one topological recursion, producing two distinct higher-genus completions of the Hurwitz-space Frobenius manifold M_{1;1}.","tokens_in":30467,"tokens_out":8990,"duration_ms":94351,"significance":"If the main structural theorem holds, it establishes a general and useful mechanism: finite-group restriction is a structure-preserving operation on the DR/DZ construction for semisimple CohFTs, and the reduced objects inherit the full Hamiltonian and tau-symmetry package. This is a substantial contribution to the DR/DZ programme, connecting the ADE/BCFG folding idea with the recent strong DR/DZ equivalence. The Section 4 argument is credible and non-circular: the reduced structures are obtained by applying ι_* to ambient objects, and the key reduction steps use Lemmas 4.2, 4.7, 4.10 together with independent results [3,5,9]. The application to M_{1;1} is striking: if correct, it gives two canonically constructed, inequivalent integrable hierarchies with the same genus-zero data. However, the advertised equivalence to genus-one topological recursion currently rests on two unverified claims, so the significance of the example is not yet fully established.","major_comments":[{"comment":"The central-invariant computation is not displayed. The text says 'One can verify similarly that the central invariants in this case are given by {1/12,1/12,1/12}', but no P_1^{Z_2×Z_2}, P_2^{Z_2×Z_2} expansions, canonical coordinates, or pencil data are given. Moreover, the effect of the rescalings 2^{1-g}Λ and ε→ε/2 on the central-invariant formula of §5.1 must be tracked explicitly. Without this computation, Corollary 5.4 and the claimed inequivalence rest on an unverified assertion.","section":"§5.2.2, after Eq. (5.7)"},{"comment":"The identification with topological recursion is asserted through an uncited rigidity statement: 'The Virasoro constraints force them to coincide up to an automorphism...' This is a nontrivial uniqueness theorem for tau functions with fixed dispersionless limit and standard Virasoro symmetries. It is neither proved nor referenced, and it is unclear whether it applies to numerically partial CohFTs that fail the loop-gluing axiom (C4). This claim is load-bearing for Theorem 1.2(2); a proof or precise citation is required.","section":"§5.2.2, Virasoro uniqueness claim"},{"comment":"Two transfer claims are stated without proof: that the flow-commutativity proof of [4] 'does not involve (C4) and remains valid if the axiom (C3) is replaced by (3.12)', and that the proof of [P_1^DR,P_2^DR]=0 in [10] only uses (|φ_α|+|φ_β|−d)η^{αβ}=0. The reduced object Λ^Γ is only a numerically partial CohFT, so these assertions are not immediate. If they are needed for the definition of the reduced DR hierarchy, they require proof; if Theorem 4.3 provides an independent derivation of the reduced bihamiltonian structure, the text should state this explicitly and remove the unproved transfer claims from the logical path.","section":"§3.4"},{"comment":"The proof of Lemma 4.2 ends with the sentence 'As for the term 1/2 θ_{α''} fP_i^{DR,α''β''}(θ_{β''}), it depends quadratically in odd variables θ^s_{α''} with index α'' in J.' This is not a complete verification of condition (2.11) for that term. The intended argument presumably uses that any derivative leaving a J-index odd variable vanishes after ι_*, but this should be written out. Since Lemma 4.2 is a central technical tool, the gap should be filled.","section":"§4.1, Lemma 4.2"}],"minor_comments":[{"comment":"The sentence 'which coincides with F^{S_4} where Q=Q^2' uses the same letter Q for two different Novikov variables. Please introduce distinct notation to avoid confusion.","section":"§5.2.2"},{"comment":"The displayed G-function '−log[(t^2)^{1/8} η(Qe^{t^3})]' appears to contain a possible typo or a missing normalization; please verify the formula and define all conventions before comparing it with −1/2 log η(Qe^{t^3}).","section":"§5.2.2, Corollary 5.5"},{"comment":"The notation 'eustr,α,s', 'A^{wk,t}_eu', and related superscripts is heavy and at times hard to parse. Defining these objects once in a glossary or local table would improve readability.","section":"§2.2 and §4.1"},{"comment":"Definition 3.9 should explicitly state which results for the DR hierarchy in the numerically partial setting are being assumed from [4] and [10] and which are proved in this paper. Currently the reader must infer this from a sentence in the introduction.","section":"§3.4"}],"recommendation":"major_revision","confidential_remarks":"The structural Theorem 1.1 appears to be the strongest and most defensible part of the paper; the Section 4 strategy via ι_* is convincing. The main obstacle to acceptance is the application: the central-invariant computation for the Z_2×Z_2 reduction and the Virasoro uniqueness step are both load-bearing and currently unverified. I would ask the authors either to provide the full computation and a citation/proof for the rigidity statement, or to state the topological-recursion equivalence as a conditional/speculative result. The §3.4 transfer assertions should also be cleaned up: they are either proved or explicitly bypassed by Theorem 4.3."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear [Colleague],\n\nThe short version: Theorem 1.1 is a genuine, probably correct extension of the BCFG-folding idea to the full DR/DZ machinery, and the proof in Section 4 is more robust than the reviewer's flagged concern about §3.4 suggests. The soft spots are mostly in Section 5.2.2, where the advertised consequence depends on computations and a uniqueness claim that are not actually shown.\n\nWhat is new: the paper proves that finite-group reduction of a semisimple CohFT's DR/DZ hierarchies preserves the bihamiltonian structure and tau structure, via a numerically-partial-CohFT formalism and a Γ-equivariance result for the strong DR/DZ equivalence. That goes cleanly beyond the ADE/BCFG case in [43]. The proof strategy is coherent: define the reduced objects as ι_* of the ambient ones, then use Lemmas 4.2 and the surjectivity of ι_* to descend the bihamiltonian and tau properties. The stress-test is right that the §3.4 transfer claims about general numerically partial CohFTs are not load-bearing for this descent; the reduced hierarchy inherits commutativity from the ambient one, even if the paper's framing makes it look otherwise. That is a minor flaw in presentation, not a flaw in the main argument.\n\nThe real soft spots are in the example. The central invariants in §5.2.2 are asserted without displaying the P1, P2 expansions or the effect of the rescalings. And the identification with genus-one topological recursion relies on an uncited rigidity statement: that Virasoro constraints plus the dispersionless limit force the tau function to coincide with the topological-recursion one, up to automorphism. That is a nontrivial uniqueness claim, and it is not obvious it holds for numerically partial CohFTs lacking (C4). The g-reduction check in Lemma 5.3 is also sketched rather than proved, though that is less concerning.\n\nThe paper is honest: it flags the deferred Virasoro-constraint analysis and the modified Virasoro constraints. The structural theorem deserves serious attention; the example needs the missing verifications before I would fully trust Theorem 1.2(2).\n\nRecommendation: send it to peer review. A good referee should ask for the central-invariant computation and a reference or proof of the Virasoro uniqueness step, but the main result is solid enough to be worth that effort.\n\nBest,\n[You]","headline":"A substantive, likely-correct generalization of BCFG folding to the DR/DZ setting, with an interesting but under-verified application in Section 5.2.2.","tokens_in":31116,"tokens_out":3210,"would_cite":true,"duration_ms":36727,"reading_group":"maybe","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":"Finite-group reduction of the DR/DZ hierarchies preserves their bihamiltonian structure and tau structure, and the reduced DR and DZ hierarchies are related by a Miura-type transformation, as shown by applying this to the orbifold Gromov–Wi","keywords":["finite group reduction","DR/DZ hierarchies","bihamiltonian structure","tau structure","cohomological field theory","orbifold Gromov-Witten theory","Hurwitz space","topological recursion"],"falsifier":"Compute the lowest-order nontrivial term in the DR flow commutator for a concrete numerically partial CohFT (satisfying C1, C2, C5, and (3.12) but not C4); if any such term fails to vanish, the weakening of the DR construction used in §3.4 is false. Alternatively, compute the central invariants of the Z2×Z2-reduced hierarchy directly from the formulas in §5.2.2 and check whether they equal {1/12,1/12,1/12}.","tokens_in":29912,"feed_emoji":"🌀","tokens_out":10659,"duration_ms":100246,"temperature":0.7,"pith_summary":"The paper establishes that finite-group reduction is a structure-preserving operation for the two integrable hierarchies associated to any semisimple cohomological field theory: the DR hierarchy and the DZ hierarchy. When a finite group acts on the cohomology, restricting to the invariant flows yields a new hierarchy that still carries a bihamiltonian structure and a tau structure, and the DR/DZ equivalence survives the reduction. The authors exhibit two finite group actions on the orbifold Gromov–Witten theory of P^1_{2,2,2,2} that produce two reduced hierarchies with the same genus-zero data—the Frobenius manifold of the Hurwitz space M_{1;1}—but genuinely different higher-genus completions. One of them is equivalent to genus-one topological recursion after a rescaling; the other has unequal central invariants. The upshot is that a single Frobenius manifold can support inequivalent higher-genus reconstructions.","feed_headline":"Reduction to fixed sector keeps DR/DZ bihamiltonian","feed_subtitle":"Restriction to invariant flows preserves DR/DZ bihamiltonian structure; one orbifold yields two distinct completions.","key_machinery":"The key object is the fixed-sector restriction: the ambient CohFT Λ is restricted to the Γ-invariant subspace of cohomology, producing Λ^Γ, which is a numerically partial CohFT—satisfying all CohFT axioms except the loop-gluing axiom, and with tree-gluing holding only numerically. The transfer is carried by Γ-equivariance of the DR/DZ machinery: the string equation is used to lift the Γ-action to the jet spaces, the DR/DZ Miura-type transformation is shown to be Γ-linearized, and an averaging lemma (Lemma 2.10) ensures that differential-polynomial quantities with a moving-sector leg vanish, which lets the Schouten–Nijenhuis bracket descend to the fixed locus.","core_discovery":"The central theorem is that finite-group reduction is structure-preserving for the DR/DZ construction. Given a semisimple CohFT with a numerical finite symmetry Γ, the restriction of the CohFT to the Γ-fixed sector is only a numerically partial CohFT—it satisfies the tree-gluing axiom numerically and fails the loop-gluing axiom—yet the associated reduced DR hierarchy still admits the bihamiltonian pair (P_1^{DR,Γ}, P_2^{DR,Γ}) and the reduced DZ hierarchy admits a compatible tau structure. Moreover, the known Miura-type equivalence between the ambient DR and DZ hierarchies restricts to a Miura-type transformation between the reduced hierarchies. In the example, applying this to the natural S","pith_inferences":["A natural extension the paper leaves implicit: if finite-group reduction is structure-preserving generally, then the classic folding constructions of Lie-algebra hierarchies (BCFG from ADE) can be obtained by DR/DZ reduction from suitable semisimple CohFTs, and the central invariants of the folded hierarchies should match those computed here.","Because the same Frobenius manifold accepts two inequivalent higher-genus completions, the genus-zero data alone do not determine the higher-genus structure; a concrete test is whether any other finite subgroup of the S4 symmetry on P^1_{2,2,2,2} yields a third set of central invariants.","The equivalence with genus-one topological recursion after rescaling ε suggests that changing the genus-expansion parameter may be the right notion of equivalence between reduced hierarchies and topological recursion; if so, one could test whether higher-genus spectral curves align with reductions of DR/DZ hierarchies under the same rescaling."],"forward_implications":["Finite-group reduction is a structure-preserving operation on the DR/DZ hierarchies for every semisimple CohFT with a numerical finite symmetry, so reduced hierarchies inherit bihamiltonian structure, tau structure, and a DR/DZ equivalence.","The genus-zero Frobenius manifold of the Hurwitz space M_{1;1} admits at least two inequivalent higher-genus integrable completions, distinguished by their central invariants.","The Z2×Z2-reduced hierarchy is, after rescaling, equivalent to the genus-one topological recursion, giving an explicit bridge between DR/DZ hierarchies and topological recursion.","The associated numerical partial CohFTs satisfy modified Virasoro constraints rather than the standard ones, so the standard linearization criterion (central invariants all 1/24) is not necessary for a tau function of topological-recursion type."],"fun_headline_variants":["Finite-group reduction preserves DR/DZ bihamiltonian structure","Reduced DR/DZ hierarchies keep tau structure","Group reduction maintains DR/DZ integrability","Orbifold reduction yields two distinct integrable hierarchies","DR/DZ reduction: bihamiltonian and tau preserved"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is that the DR-hierarchy machinery—commutativity of flows and the second Hamiltonian structure—remains valid when the CohFT axioms are weakened to the numerically partial ones satisfied by the Γ-fixed restriction; the paper refers to earlier proofs for this but does not reproduce them, and if that transfer fails the reduction theorem collapses.","fun_headline_variants_meta":{"raw":{"variants":["Finite-group reduction preserves DR/DZ bihamiltonian structure","Reduced DR/DZ hierarchies keep tau structure","Group reduction maintains DR/DZ integrability","Orbifold reduction yields two distinct integrable hierarchies","DR/DZ reduction: bihamiltonian and tau preserved"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000357,"raw_usage":{"total_tokens":1717,"prompt_tokens":633,"completion_tokens":1084,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":377,"completion_tokens_details":{"reasoning_tokens":1005}},"tokens_in":377,"tokens_out":1084,"duration_ms":10259,"temperature":1.0,"reasoning_tokens":1005,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T20:30:57.034025+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the lowest-order nontrivial term in the DR flow commutator for a concrete numerically partial CohFT (satisfying C1, C2, C5, and (3.12) but not C4); if any such term fails to vanish, the weakening of the DR construction used in §3.4 is false. Alternatively, compute the central invariants of the Z2×Z2-reduced hierarchy directly from the formulas in §5.2.2 and check whether they equal {1/12,1/12,1/12}.","supporting_citations":[],"review_version":1}