{"id":"6207c6fa-d4cc-4fb1-bb11-346ab6ec2edf","arxiv_id":"2507.05892","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For finite p-groups over commutative Noetherian rings, the Balmer spectrum of integral permutation modules decomposes over residue fields and, for elementary abelian groups, forms a Dirac scheme.","lead":"This paper describes the tensor-triangular spectrum of permutation modules for finite groups over commutative Noetherian rings, extending a theory previously known only over fields. It shows the spectrum is built from field-case fibers and, for elementary abelian p-groups, gives it the structure of a Dirac scheme.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Convention 8.1 excludes p-zero-divisor rings like Z/p^2; the omitted torsion maps c_N may change the twisted cohomology ring, so the 'technical' restriction is load-bearing and needs testing.","rationale":"The stress-test pass reviewed the main proof route leading to Theorem 1.5, including the Koszul object generation (Corollaries 4.13 and 4.16, Lemma 5.10, Theorem 5.17, Proposition 5.18), the injectivity of the comparison map (Lemma 9.3), and the Dirac scheme construction (Section 10). One apparent concern about Lemma 4.12 was resolved: since the thick tensor ideal generated by the monoidal unit in a tensor category is the whole category (the ideal is closed under tensoring with arbitrary objects, so 1⊗x ≅ x for all x), condition (2) of Lemma 4.12 is automatic for fkos objects with x_0 = 1. No clear internal inconsistency was found in the proof for rings satisfying Convention 8.1, and the main theorem appears coherent for its stated hypotheses. The remaining load-bearing concern matches the reader's weakest assumption: Convention 8.1 excludes rings where p is a nonzero zero divisor, such as Z/p^2. The excluded cases introduce new invariant elements in augmentation kernels (Remark 8.5), producing additional twisted-cohomology generators that the paper does not handle. The authors call the restriction technical, but the proof's structure changes at exactly these rings, so the claim that the restriction is not conceptual is unsupported. A direct computation for G=C_p and R=Z/p^2 would settle whether the open immersion and Dirac scheme structures persist; until such a check is done, the verdict should remain conditional.","tokens_in":37035,"tokens_out":46198,"duration_ms":511588,"concrete_test":"Fix an odd prime p and set R=Z/p^2, G=C_p. Compute H^{•,•}(C_p,R) from Definition 8.10 without imposing Convention 8.1: the maps a_N and b_N are defined as in case (C4), and check whether the invariant element p·(1+σ+...+σ^{p-1}) in the augmentation kernel gives a nonzero map c_{N,R}: 1 -> u_N[-1]. Then compute Spech(H^{•,•}), the comparison map Comp: Spc(K(C_p,R)) -> Spech(H^{•,•}), and the cover {U_E(H)} from Definitions 10.3/10.4, and determine whether Comp is injective, open, and yields a Dirac scheme structure. If the extra generator changes the spectrum or breaks the open immersion, Convention 8.1 is essential; if the same theorem holds with the extra generator, the restriction is unnecessarily narrow and the technical claim in Remark 8.2 is correct.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The main theorem (Theorem 1.5 / Corollary 10.12) is stated for Noetherian R with ann_R(p)=R or ann_R(p)=0, imposed in Convention 8.1 and labeled 'technical rather than conceptual' (Remark 8.2). This restriction is not merely simplifying: the construction of the twisted cohomology ring and its polynomial generation (Lemma 8.8) rely on the distinction between p being trivial in R and p being a non-zero-divisor. In the non-trivial case, Remark 8.5 rules out the maps c_N by asserting there is no non-trivial invariant element in the augmentation kernel. That assertion uses that p is not a zero divisor: for a non-zero-divisor p, the element p·(1+σ+...+σ^{p-1}) has coefficient sum p^2, which is nonzero in R. But for R=Z/p^2 (p>2), the element p·(1+σ+...+σ^{p-1}) is invariant and lies in the augmentation kernel, so a new map c_{N,R}: 1 -> u_N[-1] exists. The twisted cohomology ring over Z/p^2 therefore has additional generators and relations that the paper's framework does not cover. Consequently, the Dirac scheme structure and the open immersion statement are not proven for rings with p-torsion, and the paper gives no evidence that they remain true once these torsion generators are included. This is a concrete scope limitation of the central claim, directly tied to the excluded cases.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper extends Balmer–Gallauer's tensor-triangular geometry of permutation modules from a field to an arbitrary commutative Noetherian base ring R. For a finite p-group G, the authors give a set-theoretic description of Spc(K(G,R)) in terms of residue-field fibers (Theorem 5.17, Proposition 5.18), and a reduction of the topology to elementary abelian p-sections (Theorem 6.9). Under Convention 8.1, which forces p to be either 0 or a non-zero-divisor in R, they construct a twisted integral cohomology ring H^{•,•}(G,R) and prove for elementary abelian E that the comparison map Spc(K(E,R)) -> Spech(H^{•,•}(E,R)) is an open immersion and that Spc(K(E,R)) carries a Dirac scheme structure (Theorem 1.5, Corollaries 10.12 and 10.13). The paper also computes explicit spectra for cyclic groups over Z.","tokens_in":37346,"tokens_out":4631,"duration_ms":51472,"significance":"If correct, the main results form a significant extension of the known field-case tt-geometry of permutation modules to integral bases, with a concrete description of the spectrum as a set and a Dirac scheme structure for elementary abelian groups. The reduction theorem and the worked examples over Z are valuable and demonstrate the scope of the methods. The paper is well structured and builds on a clear network of prior work, including the companion paper [Gom25]. However, the central topological theorems are established only under Convention 8.1, and the paper's assertion that this restriction is purely technical is not substantiated; in fact, the restriction is load-bearing for the construction of twisted cohomology.","major_comments":[{"comment":"The exclusion of p-zero-divisor rings is not merely technical. For R = Z/p^2 with p odd, the element p·(1+σ+...+σ^{p-1}) is invariant and lies in the augmentation kernel, so a non-zero map c_{N,R}: 1 -> u_N[-1] exists even though p is nonzero in R. This contradicts the assertion in Remark 8.5 that no such map exists when p is nonzero, and it shows that the polynomial-generation statement of Lemma 8.8 fails in the presence of p-torsion. Consequently, the finite generation of H^{•,•}(G,R) (Corollary 8.13), the power-surjectivity in Corollary 8.20, and ultimately the open-immersion and Dirac-scheme conclusions (Theorem 1.5, Corollaries 10.12 and 10.13) are not proven for rings with ann_R(p) ≠ 0 whenever p is nonzero. The authors must either extend the construction to include the torsion maps or substantially revise the statement of the main theorem and remove the claim in Remark 8.2 that the restriction is technical rather than conceptual.","section":"§8, Remark 8.5 and Lemma 8.8"},{"comment":"The proof of End-finiteness in case (C4) invokes maps c_N and a decomposition involving c_{N_k}, but in case (C4) the maps c_N are not defined (Lemma 8.8 includes c_i only in case (C3)). This is an internal inconsistency in a load-bearing step: the finite-generation of Hom*_{L(H,E)}(R(E/K),1) is essential for Theorem 10.9 and hence for the Dirac scheme structure. The proof must be corrected to use only the available maps a_N and b_N, or must explain why the apparent c_N terms are harmless. As written, the argument cannot be verified by the reader.","section":"§10, Proposition 10.7, case (C4)"},{"comment":"The construction of the p=2 Koszul objects is only sketched, with the main work delegated to the proof of [BG23a, Theorem 3.1]. Since this lemma is used in Corollary 4.13 and Corollary 4.16, which in turn feed into the set-theoretic description of Theorem 5.17, a fully detailed proof should be provided locally, or at least a precise statement of which steps in [BG23a] correspond to the claims made here. The current 'essentially contained' formulation leaves too much room for ambiguity, especially in the inductive sign-modification step that must preserve the property of being a complex of permutation modules.","section":"§4, Lemma 4.7"}],"minor_comments":[{"comment":"In the proof, the reference to 'Part (4)' is incorrect; the theorem has only three numbered parts (1)–(3), and the intended reference is to part (3).","section":"§5, Theorem 5.17"},{"comment":"The notation 'Kac(G,R)' is not defined and presumably is a typo for the ideal ker(Res^G_1). Please clarify.","section":"§4, Corollary 4.13"},{"comment":"The statement that H^{•,•}(Cp,Z) = Z[a_N,b_N]/(p·a_N, p·b_N) is plausible, but the degrees of the generators should be stated consistently with Definition 8.12; the current sentence mixes the cases (C3) and (C4) without explaining which convention applies for Z.","section":"§8, Example 8.17"},{"comment":"The notation SH ⊂ H^{•,•}(G,R) is used, but the definition of the multiplicative subset omits the condition that the generators are homogeneous; please specify that SH is generated by the indicated homogeneous elements to avoid ambiguity in the localization.","section":"§10, Definition 10.4"}],"recommendation":"major_revision","confidential_remarks":"The paper depends heavily on the companion preprint [Gom25] by the second author, and several load-bearing results are quoted from it without proof. This is acceptable in principle, but the reliance is substantial and the refereeing process would be easier if the key statements from [Gom25] were summarized. The main mathematical issue is the Convention 8.1 restriction: the paper presents it as harmless, but the counterexample discussed in the report shows it changes the twisted cohomology ring. If the authors cannot extend the results to p-torsion rings, they should clearly restrict the main theorem and revise Remark 8.2. The paper is otherwise a substantial and interesting contribution, and there is no evidence of circularity beyond the explicit use of prior work."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper does what it says: it pushes the Balmer-Gallauer program for permutation modules from fields to commutative Noetherian bases. The set-theoretic description of Spc(K(G,R)) via residue fields (Theorem 5.17) and the colimit reduction to elementary abelian sections (Theorem 6.9) are genuinely new and carefully argued. The Dirac scheme result for elementary abelian p-groups over integral bases is also new, and the twisted cohomology machinery is a reasonable adaptation of the field-case framework.\n\nThe main soft spot is Convention 8.1. The authors exclude rings where p is a zero divisor (e.g., Z/p^2) and call the restriction technical. The stress-test note is right: for Z/p^2, the element p·(1+σ+...+σ^{p-1}) is an invariant in the augmentation kernel, giving a new map c_N that Remark 8.5 incorrectly rules out. So the twisted cohomology ring in excluded cases has extra generators, and the Dirac scheme statement is not merely unproven—it likely needs modification. This is a real scope limitation, though the paper is honest about it, and the motivating case R=Z is covered.\n\nA second soft spot is Lemma 4.7 (the p=2 Koszul construction), which is deferred to [BG23a] with only a sketch. The sketch is plausible, but the referee should verify that the sign modification really lands in permutation complexes, especially the interaction with Lemma 4.5.\n\nOverall, the central claims look correct for the stated cases. The p-torsion restriction is the main thing to flag; if the authors can prove something for Z/p^2 or explain why the extra torsion classes do not affect the spectrum, the paper would be stronger. As it stands, I would send this to a serious referee and accept it conditionally on those points being clarified.","headline":"Extends Balmer-Gallauer to integral base rings with real new results, but Convention 8.1 may be a bigger constraint than the authors admit.","tokens_in":37869,"tokens_out":2551,"would_cite":true,"duration_ms":30970,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["20C10","18F99","20J06","18G90","18G35"],"pacs":[],"model":"deepseek-v4-flash","headline":"For elementary abelian $p$-groups over rings like $\\mathbb{Z}$, the Balmer spectrum of integral permutation modules is an open subscheme of the homogeneous spectrum of a twisted cohomology ring and carries an explicit Dirac scheme…","keywords":["tensor-triangular geometry","Balmer spectrum","permutation modules","integral representations","twisted cohomology","Dirac scheme","elementary abelian p-groups","Koszul objects"],"falsifier":"Compute the twisted cohomology ring $H^{\\bullet,\\bullet}(E,R)$ and the comparison map for $E = C_2 \\times C_2$ and $R = \\mathbb{Z}$, and list the points of $\\operatorname{Spech}(H^{\\bullet,\\bullet}(E,\\mathbb{Z}))$; if any homogeneous prime is not hit by $\\mathrm{Comp}_{E,\\mathbb{Z}}$, the open-immersion claim is false. A sharper check: verify that the maps $\\operatorname{Spech}(H^{\\bullet,\\bullet}(E,k(\\mathfrak{p}))) \\to \\operatorname{Spech}(H^{\\bullet,\\bullet}(E,R) \\otimes_R k(\\mathfrak{p}))$ are injective for every $\\mathfrak{p} \\in \\operatorname{Spec}(R)$, since this injectivity is what the proof of Lemma 9.3 depends on.","tokens_in":36801,"feed_emoji":"🧩","tokens_out":16760,"duration_ms":154442,"temperature":0.7,"pith_summary":"The paper is trying to establish a complete description of the Balmer spectrum of the small derived category of permutation modules for a finite group over a commutative Noetherian ring, going beyond the previously understood field case. It shows the spectrum is, as a set, a disjoint union of spectra over the residue fields of the base ring, and that its topology is governed by the elementary abelian $p$-sections of the group. For elementary abelian $p$-groups, and for base rings in which the prime $p$ is either zero or a non-zero-divisor (such as the integers), it proves the comparison map into the homogeneous spectrum of the twisted cohomology ring is an open immersion and that the spectrum is a Dirac scheme. If correct, this yields explicit, computable spectra over the integers, including complete pictures for cyclic $p$-groups.","feed_headline":"Integral permutation spectra reduce to twisted cohomology","feed_subtitle":"For elementary abelian p-groups over Z, the spectrum's topology is an open immersion into a graded endomorphism ring.","key_machinery":"The load-bearing objects are the twisted cohomology ring $H^{\\bullet,\\bullet}(G,R)$, the multigraded ring of endomorphisms of tensor powers of the tensor-invertible complexes $u_{N,R}$ attached to index-$p$ normal subgroups, and the comparison map $\\mathrm{Comp}$ sending a prime to the ideal of maps whose cones lie outside it. Around these, the paper builds Koszul objects $\\mathrm{fkos}_G(H,R)$ that generate the kernels of restriction functors, and the triangular $H$-fixed points maps that transfer information between spectra of subquotients and of $G$. The final Dirac scheme structure is carried by the open cover $\\{U_E(H)\\}$ of $\\mathrm{Spc}(K(E,R))$, on which the local graded ring $\\mathcal{O}^\\bullet_E(H)$ is identified with a twist-zero localization of the twisted cohomology ring; the fact that these local rings are Noetherian and End-finite is what upgrades the injective comparison map to a homeomorphism on each open.","core_discovery":"On the paper's own terms, the central discovery is that integral permutation-module spectra are assembled from two layers: a modular fiber over the prime $p$ of $\\mathbb{Z}$ and an ordinary fiber homeomorphic to $\\operatorname{Spec}(R[1/p])$. Every prime of $K(G,R)$ is shown to be of the form $P(H,\\mathfrak{a},\\mathfrak{p})$, where $H$ is a subgroup, $\\mathfrak{p}$ is a prime of $R$, and $\\mathfrak{a}$ is a cohomological prime over the residue field $k(\\mathfrak{p})$, with explicit rules for when two such primes coincide. The topology reduces, by a homeomorphism, to a colimit over the category of elementary abelian $p$-sections of $G$. When $E$ is elementary abelian and $R$ satisfies $\\operatorname{ann}_R(p)=R$ or $\\operatorname{ann}_R(p)=0$, the comparison map $\\mathrm{Spc}(K(E,R)) \\to \\mathrm{Spech}(H^{\\bullet,\\bullet}(E,R))$ is an open immersion, and the pair $(\\mathrm{Spc}(K(E,R)), \\mathcal{O}^\\bullet_E)$ is a Dirac scheme, so the homogeneous spectrum of twisted cohomology together with an explicit open cover gives a full description of the topology.","pith_inferences":["Because the paper labels Convention 8.1 as technical, a natural test is whether the open-immersion and Dirac-scheme conclusions survive for rings where $p$ is a zero divisor, such as $\\mathbb{Z}/p^2$; the natural place to look is the twisted cohomology ring over $\\mathbb{Z}/p^2$, whose relations should differ from the $\\mathbb{Z}$ case.","The set-theoretic decomposition over residue fields suggests viewing $\\mathrm{Spc}(K(G,R))$ as a fibered space over $\\operatorname{Spec}(R)$; the paper computes the fibers, so the remaining global question is precisely which specializations cross from the ordinary fiber to the modular fiber, and the cyclic-group examples show these are controlled by the cohomological open part.","The Dirac scheme structure may allow invariants usually defined for schemes, such as the Picard group of the structure sheaf or the endomorphism ring of the unit, to be computed from the twisted cohomology ring, which would give new invariants of permutation-module categories over $\\mathbb{Z}$.","The $p=2$ Koszul construction via sign-modifications is the first place where the integral case diverges from the field case; one could check whether the same sign-modifications are needed for rings like $\\mathbb{Z}/4$, where the failure of the convention might produce genuinely new primes."],"forward_implications":["For any finite $p$-group $G$ over a Noetherian ring $R$, the Balmer spectrum of $K(G,R)$ is completely understood as a set: its points are exactly the $P(H,\\mathfrak{a},\\mathfrak{p})$, with equality governed by $G$-conjugation, residue-field behaviour, and the cohomological prime $\\mathfrak{a}$.","The topology of $\\mathrm{Spc}(K(G,R))$ is reduced to the elementary abelian case by the homeomorphism $\\varphi\\colon \\mathrm{colim}_{(H,K)\\in E(G)^{\\mathrm{op}}} \\mathrm{Spc}(K(H/K,R)) \\to \\mathrm{Spc}(K(G,R))$; for arbitrary finite groups the same argument gives a reduction through orbit categories of $p$-subgroups.","For elementary abelian $E$ and rings satisfying the convention, the comparison map is an open immersion, so the homogeneous spectrum of $H^{\\bullet,\\bullet}(E,R)$ is an affine cover of the Balmer spectrum; in particular the spectrum is a Dirac scheme.","Over $R=\\mathbb{Z}$, cyclic $p$-groups have spectra of the form described in Proposition 11.9: the ordinary fiber $\\operatorname{Spec}(\\mathbb{Z}[1/p])$ specializes into the modular fiber, which is the known field-case picture; products like $C_{p_1}\\times C_{p_2}$ glue two such pictures over a shared ordinary fiber.","The finiteness results ($H^{\\bullet,\\bullet}(E,R)$ is Noetherian, the local categories $L(H,R)$ are End-finite) make the Dirac scheme structure computable in practice, at least for small groups."],"supporting_citations":[{"why":"supplies the field-case tt-geometry, the modular fixed-point maps, and the twisted cohomology construction that this paper extends to Noetherian bases","marker":"[BG23b]"},{"why":"establishes Noetherianity of Spc(K(G,R)) and the jointly conservative residue-field base-change functors that drive the fiberwise arguments","marker":"[G´ om25]"},{"why":"provides the finite permutation resolutions and sign-modification techniques used to construct Koszul objects for p = 2","marker":"[BG23a]"},{"why":"provides the comparison map from the Balmer spectrum to the homogeneous spectrum of the endomorphism ring of the unit","marker":"[Bal10]"},{"why":"gives the homeomorphism between Spc(Db(G,R)) and Spech(H*(G,R)) and the End-finiteness lemma that upgrades injections to homeomorphisms","marker":"[Lau23]"},{"why":"supplies the exact Frobenius structure on RG-lattices and the lattice-theoretic background used in Sections 7 and 8","marker":"[BBI+23]"},{"why":"provides the definition of Dirac scheme that the paper realizes on Spc(K(E,R))","marker":"[HP23]"}],"fun_headline_variants":["Permutation module spectra over Z: a Dirac scheme twist","Integral permutation spectra: open immersion into graded ring","Balmer spectra of integral permutation modules simplified","Twisted cohomology realizes integral permutation spectra","Elementary abelian p-groups: Dirac scheme from permutation modules"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole Dirac-scheme and open-immersion result rests on the convention that, whenever the prime $p$ is not zero in the base ring $R$, it is also not a zero divisor; for rings like $\\mathbb{Z}/p^2$ this assumption fails and the main theorem is not proved.","fun_headline_variants_meta":{"raw":{"variants":["Permutation module spectra over Z: a Dirac scheme twist","Integral permutation spectra: open immersion into graded ring","Balmer spectra of integral permutation modules simplified","Twisted cohomology realizes integral permutation spectra","Elementary abelian p-groups: Dirac scheme from permutation modules"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000412,"raw_usage":{"total_tokens":2113,"prompt_tokens":905,"completion_tokens":1208,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":521,"completion_tokens_details":{"reasoning_tokens":1131}},"tokens_in":521,"tokens_out":1208,"duration_ms":9613,"temperature":1.0,"reasoning_tokens":1131,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T19:14:52.880406+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the twisted cohomology ring $H^{\\bullet,\\bullet}(E,R)$ and the comparison map for $E = C_2 \\times C_2$ and $R = \\mathbb{Z}$, and list the points of $\\operatorname{Spech}(H^{\\bullet,\\bullet}(E,\\mathbb{Z}))$; if any homogeneous prime is not hit by $\\mathrm{Comp}_{E,\\mathbb{Z}}$, the open-immersion claim is false. A sharper check: verify that the maps $\\operatorname{Spech}(H^{\\bullet,\\bullet}(E,k(\\mathfrak{p}))) \\to \\operatorname{Spech}(H^{\\bullet,\\bullet}(E,R) \\otimes_R k(\\mathfrak{p}))$ are injective for every $\\mathfrak{p} \\in \\operatorname{Spec}(R)$, since this injectivity is what the proof of Lemma 9.3 depends on.","supporting_citations":[],"review_version":1}