{"id":"258949d3-f5b0-4552-a9e8-ea6451ba49b7","arxiv_id":"2507.11478","paper_version":2,"verdict":"CONDITIONAL","confidence":"LOW","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For odd genus g, the integral Chow rings of the balanced hyperelliptic Prym component and of the unordered divisor stack are explicitly computed as quotient rings.","lead":"Mathematicians computed the intersection ring, called the Chow ring, of the hardest component of the moduli stack of hyperelliptic Prym pairs in odd genus, and used it to describe all hyperelliptic Spin curve components. The computation completes a multi-paper program and supplies explicit algebraic presentations for these moduli spaces.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 1.7 rests entirely on [CL25, Prop. 3.5], an unproved µ2-root-gerbe formula imported from the companion paper; the paper verifies neither its hypotheses nor the resulting ring extension for a base with 2-torsion, so the central presentation is only as secure as that imported result.","rationale":"The reader's weakest-assumption analysis is correct: the central theorem is not an independent computation but an application of two imported results from [CL25]. The most load-bearing of these is Proposition 3.5, because Theorem 1.7 is obtained from Theorem 1.6 precisely by adjoining t with the single relation 2t=β1+γ. If the root-gerbe Chow-ring formula has hidden hypotheses, or if the line-bundle identification in §4.7 is off by a sign, the presentation collapses at the level of CH^1. The paper is careful and contains substantial independent work, including the CKP results in §3.3 and the long pushforward computations of §4.4–4.6; if the imported proposition is valid in this setting, the argument is coherent. But the reliance is real and unverified, and the base stack has 2-torsion, so a conditional verdict is the right level of confidence. I therefore see no reason to change the reader's CONDITIONAL verdict, and the same weak point is identified in both readings.","tokens_in":41420,"tokens_out":27378,"duration_ms":331939,"concrete_test":"Independently re-derive [CL25, Proposition 3.5] from the root-stack/projective-bundle formula and check its hypotheses for the base Y=[D_{g+1,g+1}/µ2]: in particular, verify that CH*(Y) injects into CH*(X) for X=RH_g^{(g+1)/2}, that π_*π^*=2, and that the only new degree-1 relation is 2t=β1+γ. Then compute, for the minimal case g=3, the first two graded pieces CH^1 and CH^2 of RH_3^2 by an independent method (e.g. directly from the presentation in Theorem 1.4 using equivariant Chow and a torus restriction), and compare with the truncation of the presentation in Theorem 1.7. If CH^1 is not the predicted Z/4⊕Z/2 with the stated generators, or if any additional degree-1 or degree-2 relation appears, the imported root-gerbe formula has been misapplied and the theorem needs correction.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The decisive step is in §4.7: after Theorem 1.6, the paper applies [CL25, Proposition 3.5] to the µ2-gerbe π: RH_g^{(g+1)/2} → [D_{g+1,g+1}/µ2], concluding CH*(RH_g^{(g+1)/2}) = CH*([D_{g+1,g+1}/µ2])[t]/(2t-(β1+γ)). This asserts that the gerbe adds exactly one degree-1 class t and no further relations, i.e. that the pullback π* is compatible with a single polynomial extension. The base ring has 2-torsion (2γ=0, and later 2β1=0 in I), and the paper only proves Chow-Künneth for classifying stacks (Proposition 3.13), not for the quotient stack [D_{g+1,g+1}/µ2] on which the formula is applied. The needed hypotheses of [CL25, Proposition 3.5] are not checked. The companion identification c1(L)=β1+γ is also asserted in one line; if the relevant character of G is the literal product ab rather than the determinant (-1)^ε ab, the relation would be 2t=β1, changing even CH^1. An independent low-degree computation of CH*(RH_g^{(g+1)/2}) is therefore needed before Theorem 1.7 can be regarded as established. The duplicated “Proof of Proposition 4.24” label in §4.6.3 and the apparently undefined “c1” in the first bullet of Theorem 1.6 are additional editorial red flags, though not by themselves decisive.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper computes the integral Chow rings of the components RH_g^{(g+1)/2} of the moduli stack of hyperelliptic Prym pairs for odd g ≥ 3, as well as the integral Chow ring of the auxiliary quotient stack [D_{g+1,g+1}/µ2]. The strategy is to present these stacks as quotient stacks, pass to GL3-counterparts and maximal-torus extensions, compute the ideals generated by pushforwards along a Chow envelope by the S_r, M_r, and H_r maps, and finally apply a µ2-root-gerbe formula from the authors' companion paper [CL25] to pass from [D_{g+1,g+1}/µ2] to RH_g^{(g+1)/2}. The paper also gives a geometric interpretation of the generators and, as an application, presentations of the Chow rings of all components of the moduli stack of hyperelliptic Spin curves of odd genus.","tokens_in":41764,"tokens_out":11804,"duration_ms":127885,"significance":"If correct, these are substantial new computations in integral Chow theory of moduli stacks. The paper displays a high level of technical sophistication: it develops Chow-Künneth results for several classifying stacks, uses GL3-counterparts to reduce PGL2-equivariant pushforwards to torus-equivariant ones, and carries out large explicit computations of equivariant Chow envelopes. The final presentations are explicit and come with a geometric interpretation of the generators, and the application to Spin curves is a nice consequence. The main caveat is that the final step relies on a technical result imported from the companion paper [CL25], whose hypotheses are not checked in this paper.","major_comments":[{"comment":"The proof of Theorem 1.7 applies [CL25, Proposition 3.5] to the µ2-gerbe RH_g^{(g+1)/2} → [D_{g+1,g+1}/µ2] and concludes that CH*(RH_g^{(g+1)/2}) is obtained from CH*([D_{g+1,g+1}/µ2]) by adjoining a degree-one class t with the single relation 2t = β1 + γ. The hypotheses of [CL25, Proposition 3.5] are not stated or verified in the paper. In particular, the base ring has 2-torsion (2γ = 0 and then 2β1 = 0 in I), and the base stack [D_{g+1,g+1}/µ2] is not a classifying stack, whereas the Chow-Künneth property is proved in Proposition 3.13 only for classifying stacks of the relevant groups. Additionally, the identification c1(L) = β1 + γ is asserted in a single sentence; it depends on the choice of µ2-linearization of O(1) ⊠ O(1) on P(W) × P(W), and a different choice would give 2t = β1, changing even CH^1. The authors should either state and verify the hypotheses of the imported proposition for this specific gerbe or provide an independent low-degree computation of CH*(RH_g^{(g+1)/2}) to confirm the relation.","section":"§4.7, Theorem 1.7"},{"comment":"The displayed identity \"-2M1*(1) − S1*(1) = 2β1\" is arithmetically incorrect. Substituting Lemma 4.12 (M1*(1) = −(g+1)β1) and Lemma 4.9 (S1*ϕ1*(1) = −2gβ1) gives −2M1*(1) − S1*(1) = 2(g+1)β1 + 2gβ1 = (4g+2)β1 = 2(2g+1)β1, not 2β1. The desired relation 2β1 ∈ I is recoverable because gcd(2g, g+1) = 2 for odd g, but the equation written in the proof is false. Please correct the displayed identity or replace it with the correct Bezout combination of M1*(1) and S1*(1).","section":"§4.7, proof of Theorem 1.6"}],"minor_comments":[{"comment":"The first bullet in the list of relations contains an undefined symbol \"c1\". The relations coming from B(G × PGL2) are 2γ, γ(β1 + γ), and 2c3, as used in the proof; please clarify or delete \"c1\".","section":"Theorem 1.6"},{"comment":"The label \"Proof of Proposition 4.24\" appears twice in this subsection: the first occurrence is actually a strategy outline, and the second is the real proof. Rename the first occurrence to \"Strategy\" or similar to avoid confusion.","section":"§4.6.3"}],"recommendation":"major_revision","confidential_remarks":"The manuscript depends heavily on the companion paper [CL25], especially Proposition 3.5, which is not reproved here. The editor may wish to confirm that Part I is in its final, accepted form before Part II is published. The false identity in the proof of Theorem 1.6 and the undefined \"c1\" in Theorem 1.6 should also be corrected."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First thing to know: this is a serious paper, probably correct in its main lines, but it is not self-contained. The headline result — the integral Chow ring of the balanced component RH_g^((g+1)/2) — is genuinely new; Part I left it open. The application to Spin curves is a nice bonus. The methods are the right ones: Chow envelopes, μ2-equivariant computations, GL3-counterparts, and the exposition is clear given the unavoidable technical density.\n\nThe soft spot is Theorem 1.7. The computation of CH*([D_{g+1,g+1}/μ2]) in Theorem 1.6 is a long but structured calculation, and I did not find an obvious contradiction. But Theorem 1.7 is obtained by applying [CL25, Prop. 3.5] to the μ2-root gerbe RH_g^((g+1)/2) → [D_{g+1,g+1}/μ2], and the paper does not verify the hypotheses of that proposition for a base with 2-torsion. The base ring has 2γ=0 and later 2β1=0, and the paper only proves Chow-Künneth for classifying stacks, not for the quotient stack in question. The companion formula may be fine, but it is load-bearing and unproved here. Similarly, c1(L)=β1+γ is asserted in one line; if the relevant character is the determinant rather than the literal product, the relation would be 2t=β1, changing even CH^1. An independent low-degree computation would resolve this quickly.\n\nAlso, the paper imports a great deal from Part I — not just the root-gerbe formula but also Propositions 3.4, 3.5, 3.16 and several pushforward lemmas. That is not a flaw per se, but it means a referee needs both papers in hand. Minor editorial issues: duplicated “Proof of Proposition 4.24” label in §4.6.3, and an undefined “c1” in the first bullet of Theorem 1.6. These are easy to fix.\n\nWho is this for? Specialists in the Chow rings of moduli stacks. It deserves a serious referee. I would not desk-reject it; I would send it out with a request to check the root-gerbe step and the c1(L) identification carefully. If those hold, the paper is a solid contribution.","headline":"Serious, technically strong, but Theorem 1.7 leans on an unverified root-gerbe formula from Part I; referee should check it before accepting.","tokens_in":42345,"tokens_out":5066,"would_cite":true,"duration_ms":54913,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14C15","14H10","14H40","14L30"],"pacs":[],"model":"deepseek-v4-flash","headline":"For odd genus, integral Chow rings of hyperelliptic Prym stacks are now explicit.","keywords":["integral Chow ring","moduli stacks","hyperelliptic Prym pairs","Spin curves","quotient stacks","equivariant Chow groups","root gerbes"],"falsifier":"Compute $\\mathrm{CH}^2([D_{4,4}/\\mu_2])$ for the smallest case $g=3$ by an independent method, for instance by resolving the discriminant locus directly or using the known Chow ring of the moduli of smooth genus-$3$ hyperelliptic curves, and compare the rank and torsion of the degree-$2$ group with the quotient ring in Theorem 1.6; a single mismatch would falsify the presentation.","tokens_in":41198,"feed_emoji":"🧮","tokens_out":6641,"duration_ms":79685,"temperature":0.7,"pith_summary":"This paper pins down the integral Chow ring of the deepest component of the moduli stack of hyperelliptic Prym pairs for every odd genus: the component where the two divisors recording the Prym pair have equal degree $g+1$. The authors prove that this ring is a quotient of a polynomial ring in six generators by an explicit finite list of relations, and that the same computation gives the Chow ring of the auxiliary stack of unordered pairs of disjoint degree-$(g+1)$ divisors on the projective line. A corollary is a complete description of the integral Chow rings of all irreducible components of the moduli stack of hyperelliptic Spin curves of odd genus. The payoff is concrete: the generators are Chern classes of natural vector bundles, so the presentation can be used for intersection-theoretic computations on these moduli spaces.","feed_headline":"Odd-genus Prym pair Chow rings pinned down","feed_subtitle":"Explicit generators and relations for all components of the hyperelliptic Prym stack, plus the Spin-curve components.","key_machinery":"The load-bearing device is the presentation of $\\mathcal{RH}_g^{(g+1)/2}$ as the $\\mu_2$-root gerbe over $[D_{g+1,g+1}/\\mu_2]$ associated with the pullback of $\\mathcal{O}_{\\mathbb{P}(W_{g+1})}(-1)$, whose first Chern class is $\\beta_1+\\gamma$. Combined with the companion paper's formulas for Chow rings of $\\mu_2$-root gerbes and for recovering $\\mathrm{GL}_3$-equivariant Chow ideals from their maximal-torus extensions, the problem reduces to computing the ideal generated by pushforwards along the equivariant Chow envelope maps $M_r$, $S_r$, and $H_r$ of the singular locus of pairs of forms; the geometric generators $\\beta_1,\\beta_2,\\gamma$ are the Chern classes of a natural rank-two bundle $\\widetilde{N}_{(g+1)/2}$ on $[D_{g+1,g+1}/\\mu_2]$.","core_discovery":"The paper establishes, for every odd $g \\ge 3$, the explicit presentations $\\mathrm{CH}^\\ast([D_{g+1,g+1}/\\mu_2]) \\cong \\mathbb{Z}[\\beta_1,\\beta_2,\\gamma,c_2,c_3]/I$ and $\\mathrm{CH}^\\ast(\\mathcal{RH}_g^{(g+1)/2}) \\cong \\mathbb{Z}[\\beta_1,\\beta_2,\\gamma,c_2,c_3,t]/(I + \\langle 2t-(\\beta_1+\\gamma)\\rangle)$, where $I$ is the ideal generated by the nine explicit relations listed in Theorem 1.6. The proof reduces the computation to the pushforwards of an equivariant Chow envelope of the discriminant locus of pairs of degree-$(g+1)$ forms, and the final ring for the Prym component is obtained by applying a root-gerbe formula to the $\\mu_2$-gerbe $\\mathcal{RH}_g^{(g+1)/2} \\to [D_{g+1,g+1}/\\mu_2]$. As an application, the decomposition $\\mathcal{SH}_g \\cong \\mathcal{H}_g \\sqcup \\mathcal{RH}_g$ yields presentations and Chow rings for all irreducible components of the moduli stack of hyperelliptic Spin curves of odd genus.","pith_inferences":[],"forward_implications":["For odd $g$, the integral Chow ring of $\\mathcal{RH}_g^{(g+1)/2}$ is completely and explicitly known: six generators and a finite relation list.","Together with previously known rings, the integral Chow ring of every irreducible component of the moduli stack of hyperelliptic Spin curves of odd genus is obtained.","The relation $2t=\\beta_1+\\gamma$ shows that the root-gerbe class $t$ is redundant after inverting $2$ but contributes integral $2$-torsion.","The generators $\\beta_1,\\beta_2,\\gamma$ are Chern classes of geometrically defined vector bundles, so the presentation supports concrete intersection-theoretic calculations on these stacks.","The intermediate computation of $\\mathrm{CH}^\\ast([D_{g+1,g+1}/\\mu_2])$ gives a complete description of the Chow ring of unordered pairs of disjoint divisors of the same even degree in $\\mathbb{P}^1$.","For even $g$ there is no equal-degree \"middle\" pair, so the same root-gerbe model does not apply directly; a variant tracking the complementary divisor of degree $g+2$ might handle the even-genus components.","The equivariant Chow-envelope technique for the discriminant of pairs of forms could be reused for other moduli of covers where a $\\mathrm{GL}_3$-counterpart exists.","A quick consistency test of the presentation is to reduce the relations modulo $2$ and compare the resulting additive structure with the known rational Chow ring, which should show only the $2$-torsion visible in $2t=\\beta_1+\\gamma$ and $2\\gamma=0$."],"supporting_citations":[{"why":"Supplies the two load-bearing tools: the Chow ring of a $\\mu_2$-root gerbe and the maximal-torus reduction for $\\mathrm{GL}_3$-equivariant Chow ideals, plus the first paper's computations for the other components.","marker":"[CL25]"},{"why":"Introduces $\\mathrm{GL}_3$-counterparts, computes the Chow ring of $\\mathrm{BPGL}_2$, and provides pushforward computations for multiplication and squaring maps.","marker":"[DL18]"},{"why":"Provides the foundations of equivariant intersection theory and the Chow-envelope technique used to compute the relevant pushforward ideals.","marker":"[EG98]"},{"why":"Gives the Chow ring of the classifying space of $G=(G_m\\times G_m)\\rtimes\\mu_2$ and the push-pull formulas used throughout the ideal computations.","marker":"[Lar21]"},{"why":"Provides the key lemma describing the map from Weierstrass divisors to $2$-torsion line bundles that underlies the component decomposition of $\\mathcal{RH}_g$.","marker":"[Ver13]"},{"why":"Establishes the splitting $\\mathcal{J}_g[2]\\cong\\mathcal{R}_g\\sqcup\\mathcal{M}_g$ used for the Spin-curve application.","marker":"[CIL24]"},{"why":"Gives the quotient-stack presentations of $\\mathcal{H}_g$ and $\\mathcal{D}_{2g+2}$ used as the starting point for the root-gerbe description.","marker":"[AV04]"}],"fun_headline_variants":["Odd-genus Prym Chow rings fully computed","Explicit Chow rings for hyperelliptic Prym stacks","Prym pair Chow rings for odd genus solved","Integral Chow rings for odd-genus Prym pairs","Chow rings of Prym stacks for odd genus"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The computation inherits two technical results from the authors' companion paper — the formula for the Chow ring of a $\\mu_2$-root gerbe and the reduction of $\\mathrm{GL}_3$-equivariant Chow ideals to their maximal-torus restrictions — and applies them without reproving them; if either of those results is not valid for these stacks, the stated presentation collapses.","fun_headline_variants_meta":{"raw":{"variants":["Odd-genus Prym Chow rings fully computed","Explicit Chow rings for hyperelliptic Prym stacks","Prym pair Chow rings for odd genus solved","Integral Chow rings for odd-genus Prym pairs","Chow rings of Prym stacks for odd genus"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000202,"raw_usage":{"total_tokens":1412,"prompt_tokens":1007,"completion_tokens":405,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":623,"completion_tokens_details":{"reasoning_tokens":328}},"tokens_in":623,"tokens_out":405,"duration_ms":4398,"temperature":1.0,"reasoning_tokens":328,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T17:08:06.792244+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute $\\mathrm{CH}^2([D_{4,4}/\\mu_2])$ for the smallest case $g=3$ by an independent method, for instance by resolving the discriminant locus directly or using the known Chow ring of the moduli of smooth genus-$3$ hyperelliptic curves, and compare the rank and torsion of the degree-$2$ group with the quotient ring in Theorem 1.6; a single mismatch would falsify the presentation.","supporting_citations":[],"review_version":1}