{"id":"25d74d5c-239b-43ac-8346-5c45979b8c70","arxiv_id":"2507.00309","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":3.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A survey codifying geometric modeling and two-tower methods for group actions on formal moduli stacks, with the Lubin-Tate action as the guiding example.","lead":"This survey paper lays out two ways to understand what happens when a group acts on a formal moduli stack: geometric modeling and the two-tower method. It applies these to the Lubin-Tate action, the automorphisms of a formal group acting on its deformations, which is central to modern homotopy theory and number theory.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 1.21 is the load-bearing bridge for Cor 6.3; its proof only equates underived global sections, and under that reading it is false for non-affine X, so the continuous-cohomology payoff is not established as written.","rationale":"The reader's weakest assumption was the existence of a G-equivariant deformation equivalence F. That is indeed a nontrivial input for the examples, but it is a limitation the paper acknowledges: Cor 6.1 is a conditional criterion, and Serre-Tate/Carayol supply the condition in the flagship cases. The place where the central claim is least secure technically is the cohomological bridge: without Lemma 1.21, Cor 6.3 does not follow from Cor 6.1, and the abstract's promise to 'compute cohomology related to this group action' is unfulfilled. The proof of Lemma 1.21 is an adjunction at the level of global sections and does not justify the derived equivalence. In an ∞-categorical treatment the statement may be true with F(X) read as RΓ(X,F), but the paper does not say this, and Cor 1.22's notation O(Def*_X) invites the underived reading, under which the lemma has simple counterexamples. This is a correctness risk, not a stylistic one; however it is repairable by stating the lemma in derived form and noting that the deformation stacks in the applications are affine formal schemes (or by proving the general Cartan-Leray statement). Hence my verdict agrees with the reader's CONDITIONAL: the framework is plausible and useful, but the cohomological machinery needs a precise statement and proof before the central claims can be relied upon. I do not see grounds to reject or to accept outright.","tokens_in":15929,"tokens_out":21668,"duration_ms":247849,"concrete_test":"Rewrite Lemma 1.21 with derived global sections explicit: prove RΓ(X/G,F) ≃ RΓ_cts(G,RΓ(X,F)) via the Cartan-Leray spectral sequence E_2^{p,q} = H^p_cts(G,H^q(X,F)) ⇒ H^{p+q}(X/G,F). Then test the underived statement against X = P^1_C with trivial G = Z/2 and F = O: H^1(P^1/G,O) ≠ H^1_cts(Z/2,O(P^1)); this settles whether the current text needs a correction (derived statement) or has a counterexample (underived statement).","verdict_should_be":"UNCHANGED","load_bearing_attack":"Corollary 6.3 ('group time') is the advertised cohomological payoff: it turns the stack equivalence of Cor 6.1 into H*_cts(G,O(Def*_X)) ≃ H*_cts(G,O(Def*_F(X))) by way of Lemma 1.21. Lemma 1.21 states H*(X/G,F) ≃ H*_cts(G,F(X)) for F ∈ QCoh(X/G), but its proof only passes through the adjunction for ordinary global sections: Hom(よ(X/G),F) ≃ Hom(よ(pt/G),p_*F) ≃ Hom(よ(pt),F(X)). With the natural underived reading of F(X) (as used for O(Def*_X) in Cor 1.22), the lemma is false in general: take G = Z/2 acting trivially on X = P^1_C and F = O; then H^1(X/G,O) = H^1(P^1,O) = C, while H^1_cts(G,O(X)) = H^1(Z/2,C) = 0. To make the argument work, F(X) must mean the derived global sections RΓ(X,F), and one must prove RΓ(X/G,F) ≃ RΓ_cts(G,RΓ(X,F)) (e.g., via the Cartan-Leray spectral sequence). The paper never states this derived formulation or supplies the missing proof; Cor 6.3's proof moreover cites a nonexistent Lemma 6.3. The intended affine deformation spaces (Lemma 5.30) may survive, but the general cohomological claim is not established as written.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes a framework for studying profinite group actions on formal moduli stacks, with the Lubin-Tate action as the motivating example. Two methods are developed: 'geometric modelling', which passes from a G-equivariant equivalence of deformation functors Def^*_X ≃ Def^*_{F(X)} to an equivalence of the corresponding G-deformation stacks Def^G_X ≃ Def^G_{F(X)}, and the 'two tower method', which compares quotients of a G×G'-torsor by residual actions. The central cohomological bridge is Lemma 1.21, which asserts that for a quasi-coherent sheaf F on X/G one has H^*(X/G,F) ≃ H^*_cts(G,F(X)). This is used to convert stack equivalences into equivalences of continuous group cohomology, e.g., H^*_cts(G,O(Def^*_X)) ≃ H^*_cts(G,O(Def^*_{F(X)})). Section 5 reviews formal groups and the Lubin-Tate deformation ring, and Section 6 states the resulting criteria. The paper is explicitly a survey/prolegomenon collecting known examples (Serre-Tate, Carayol, Rapoport-Zink, Barthel-Schlank-Stapleton-Weinstein) into a common language.","tokens_in":16274,"tokens_out":12606,"duration_ms":141717,"significance":"If the main claims are established, the paper would provide a useful conceptual organizing framework for the Lubin-Tate action and its appearance in chromatic homotopy theory and the Jacquet-Langlands correspondence. The core formal lemmas (2.4, 3.3, 4.1) are plausible and are essentially formal consequences of the definitions, and the examples show that the framework captures substantial known mathematics. The paper ships no machine-checked proofs and no new computational results; its value lies in synthesis and in the proposed criterion for geometric modelling. However, the advertised cohomological payoff is not rigorously established as written: Lemma 1.21 is false in the form stated, and the proof of the central 'group time' corollary cites nonexistent lemmas. These issues are local in the sense that the intended statement is likely correct for the affine deformation spaces that appear in the examples, but they must be repaired before the survey can serve as a reliable reference.","major_comments":[{"comment":"The lemma as stated is false if F(X) means ordinary global sections. For example, take X = P^1_C, G = Z/2 acting trivially, and F = O_X. Then H^1(X/G,O) ≃ H^1(P^1,O) ≃ C, while H^1_cts(G,O(X)) ≃ H^1(Z/2,C) = 0. The proof only relates underived global sections through Hom(よ(X/G),F) ≃ Hom(よ(pt),F(X)) and then asserts the derived conclusion, so it does not justify the step to H^*_cts. To make the lemma correct, either F(X) must mean the derived global sections RΓ(X,F) and one must prove RΓ(X/G,F) ≃ RΓ_cts(G,RΓ(X,F)) (e.g., via the Cartan-Leray spectral sequence), or the lemma must be restricted to X with RΓ(X,F) ≃ F(X), such as affine X. The affine restriction covers the deformation rings of Lemma 5.30, but the general statement in the abstract and Section 1.2 is not supported. This issue affects Corollary 1.22, Corollary 3.5, Corollary 4.3, Corollary 6.3, and Corollary 6.6.","section":"§1.2 (Lemma 1.21)"},{"comment":"The proofs of Corollary 6.2 and Corollary 6.3 cite 'Lemma 6.2' and 'Lemma 6.3', neither of which exists in the manuscript. The same problem appears in Corollary 3.5, which cites 'Lemma 6.2'; Corollary 4.2 cites 'Corollary 4.1'; and Corollary 6.5 cites 'Lemma 4.2'. Because Corollary 6.3 ('group time') is the advertised cohomological payoff of the paper, this is not a purely typographical issue: as written, the proof of the central claim is missing. The authors should correct the cross-references and supply a complete proof of Corollary 6.3 that explicitly invokes the (repaired) Lemma 1.21 or Corollary 1.22 together with Corollary 3.4 and Lemma 2.4.","section":"§6 (Corollaries 6.2 and 6.3)"},{"comment":"Corollary 6.1, the 'corollary of greed', is a restatement of Lemma 3.3 with the target stack N renamed M♠fg1. The question posed at the start of Section 6 is answered by 'there exists a functor F with the stated property', which is precisely the hypothesis of Lemma 3.3. If Section 6 is intended as a summary of the earlier framework, it should be labelled as such; as presented, the 'Criterion Theorem' gives no new criterion beyond the earlier lemmas. This is a presentation issue, but it directly bears on the paper's claim to 'fill a gap' and should be clarified so readers are not misled about the novelty of Section 6.","section":"§6 (Corollary 6.1)"}],"minor_comments":[{"comment":"The proof of Lemma 2.4 is very compressed and contains a garbled line: 'Consider the map from Def^{Aut_k(X)}_X → B Aut_k(X), (X -> X') ↦ (X|_k -> X)'. The intended pullback diagram is plausible, but the reader needs a clean statement of the G-action on Def^*_X and a verification that the big pullback square is indeed a G-torsor. Please rewrite this proof with enough detail to make the equivalence Def^G_X ≃ (Def^*_X)/G fully transparent.","section":"§2 (Lemma 2.4)"},{"comment":"The phrase 'co-represented by a ring' is used where the intended meaning is likely 'represented by the formal scheme Spf A' or 'co-represented by the topological ring A'. The current wording conflates a functor represented by a ring with a functor whose values are homomorphisms out of a ring; please make the convention consistent.","section":"§5.5 (Lemma 5.30, Corollary 5.32)"},{"comment":"The notation M♠fg1 is used in Section 6 before it is defined; please introduce it at the beginning of Section 6 or earlier, explicitly setting M♠fg1 := Mfg1 or Ms_fg1 and saying what '♠' is meant to convey.","section":"§6 (Notation)"},{"comment":"The proof of Corollary 1.20 asserts an identification of the Cartan-Leray spectral sequence with the standard complex computing continuous group cohomology, but gives only a brief sketch. A reference or a few more lines explaining why the differentials can be identified would make the argument self-contained.","section":"§1.2 (Corollary 1.20)"},{"comment":"The symbol よ (Yoneda) is used throughout the proof of Lemma 1.21 without being defined. Please define the Yoneda embedding at first use, or replace it with more conventional notation for a representable sheaf or stack.","section":"§1.1 (Notation)"},{"comment":"Some references are cited only by a short tag without full publication data, in particular [HM] (Heyer and Mann) and [Pst] (Pstragowski). Please complete the bibliography entries so readers can locate the sources of Lemmas 1.17, 1.19, 5.12, 5.13, 5.21, 5.22, and 5.23.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper is best read as an informal survey that collects important examples and proposes a unifying language. Its core geometric-modelling lemmas are plausible, but the central cohomological bridge is not established as written: Lemma 1.21 is false under the ordinary-global-sections reading, and the 'group time' corollaries cite nonexistent lemmas. I believe this is repairable either by proving the derived version of Lemma 1.21 or by explicitly restricting to affine deformation spaces, which are the cases actually used in the Lubin-Tate examples. I recommend major revision rather than rejection because the intended framework is valuable and the fix is local, but the present version cannot be accepted without these repairs."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is a survey, not a new theorem, but it does something useful: it collects the geometric modeling and two-tower methods for understanding profinite group actions on deformation stacks, and frames them around the Lubin-Tate action. The examples are well chosen—Serre-Tate, Carayol's Shimura varieties, Rapoport-Zink, the BSSW24 rational vanishing argument—and the condensed/proetale setup for quotient stacks is a reasonable way to discuss profinite quotients. The author is honest that the 'criterion theorem' restates earlier lemmas.\n\nThe soft spots are real and need attention. Lemma 1.21 is the bridge that turns stack equivalences into comparisons of continuous group cohomology, and its proof is too quick. It only runs the adjunction for ordinary global sections. As written, the statement is false in general: take G = Z/2 acting trivially on X = P^1 over C, with F = O. Then H^1(X/G, O) = H^1(P^1, O) = C, but H^1_cts(G, O(X)) = H^1(Z/2, C) = 0. To make the lemma true you need the derived reading: RΓ(X/G, F) ≃ RΓ_cts(G, RΓ(X, F)), which requires a Cartan-Leray spectral sequence. The paper never states or proves that. The intended deformation spaces are affine (W[[u1,...,u_{h-1}]]), so the examples might survive, but the general cohomology claim is not established.\n\nThere are also mechanical issues: Corollaries 6.2 and 6.3 cite Lemma 6.2 and Lemma 6.3, which don't exist; Corollary 6.5 cites Lemma 4.2 instead of Corollary 4.2; the proof of Lemma 2.4 has a garbled line; and there are assorted typos. These would be easy to fix.\n\nWho is this for? Someone looking for a map of the Lubin-Tate action literature, or an entry point into BSSW24-style two-tower arguments. It could serve that role after revision. I wouldn't cite it in its current form, and I'd want the Lemma 1.21 question resolved before trusting the cohomological statements. That said, the organizational value is real and the author clearly knows the material. I'd send it to a referee, but the referee should be asked to push for a correct and explicit statement of the cohomology bridge and a cleanup of the cross-references.","headline":"A useful survey of methods for the Lubin-Tate action, but the cohomology bridge (Lemma 1.21) is underproved and false as stated in general; needs revision before it can be relied on.","tokens_in":16840,"tokens_out":3310,"would_cite":false,"duration_ms":32851,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14D23","14L05","55N22"],"pacs":[],"model":"deepseek-v4-flash","headline":"A G-equivariant match between deformation problems makes quotient stacks and their cohomology agree.","keywords":["Lubin-Tate action","formal groups","deformation stacks","profinite group actions","geometric modelling","two tower method","continuous group cohomology","quotient stacks"],"falsifier":"A concrete check: take a finite subgroup $G$ of automorphisms of a supersingular elliptic curve $E$ over an algebraically closed field of characteristic $p$, and compare the $G$-invariant subring of the completed local ring of the moduli stack of elliptic curves at $E$ with $W[[u_1]]^G$ for the corresponding height-two formal group. If these invariant subrings are not isomorphic, Corollary 6.1 fails; if they are, the supersingular elliptic curve example is confirmed.","tokens_in":15681,"feed_emoji":"🧮","tokens_out":19429,"duration_ms":184784,"temperature":0.7,"pith_summary":"This paper sets up a general recipe for understanding a profinite group action on a formal moduli stack: find a different stack, equipped with a different group action, and a functor between them that identifies the two deformation problems while commuting with the group. The main criterion (Corollary 6.1) says that if such a $G$-equivariant equivalence of deformation functors exists, then the quotient stacks—and therefore all their coherent cohomology, including continuous group cohomology—are the same. The motivating case is the Lubin-Tate action, the action of the automorphism group of a one-dimensional formal group on its deformation space. A reader should care because this turns a floppy action into a computable one: the paper gives models for supersingular elliptic curves, for height $p-1$, and for general height $h$, plus a two-tower method that compares actions of different groups. If the recipe is right, hard cohomology computations for automorphism groups of formal groups reduce to invariant-theoretic computations on more geometric stacks.","feed_headline":"Swap deformation problems and keep their group cohomology","feed_subtitle":"The Lubin-Tate action can be traded for a tractable stack without changing continuous group cohomology.","key_machinery":"The load-bearing objects are the deformation groupoid $\\mathrm{Def}^G_X$—the stacky quotient of the star-deformation functor (deformations with isomorphisms reducing to the identity on the special fibre) by the profinite group $G$—together with the identification $\\mathrm{Def}^G_X \\simeq (\\mathrm{Def}^{\\star}_X)/G$ (Lemma 2.4), and the site-theoretic identity $H^*(X/G,\\mathcal{F}) \\simeq H^*_{\\mathrm{cts}}(G,\\mathcal{F}(X))$ (Lemma 1.21) that converts coherent cohomology of quotient stacks into continuous group cohomology. Around these, the paper places the moduli of one-dimensional formal groups (graded or ungraded), the height invariant that classifies formal groups over algebraically closed fields, the constant profinite automorphism group $\\operatorname{Aut}_k(F)$, and the rings $W[[u_1,\\ldots,u_{h-1}]]$ and $W[[u_1,\\ldots,u_{h-1}]][\\beta^{\\pm 1}]$ representing star-deformations. The functor $F:\\mathcal{M}\\to\\mathcal{M}^{\\natural}_{\\mathrm{fg1}}$ is the 'puppet' that carries the group action into this tractable setting.","core_discovery":"On the paper's own terms, the central discovery is that the Lubin-Tate action—and more generally any profinite group action on a formal moduli stack—can be 'modeled' by a more tractable action without losing cohomological information. Precisely, for prestacks $\\mathcal{M}$ and $\\mathcal{N}$ with a $G$-equivariant functor $F:\\mathcal{M}\\to\\mathcal{N}$, the paper proves that a $G$-equivariant equivalence of deformation functors $\\mathrm{Def}^{\\star}_X \\simeq \\mathrm{Def}^{\\star}_{F(X)}$ implies an equivalence of quotient stacks $\\mathrm{Def}^G_X \\simeq \\mathrm{Def}^G_{F(X)}$ (Lemma 3.3, Corollary 6.1). Combined with Lemma 1.21, which identifies the coherent cohomology of such a quotient stack with continuous group cohomology $H^*_{\\mathrm{cts}}(G,\\mathcal{O}(\\mathrm{Def}^{\\star}_X))$, this gives invariance of these cohomology groups under modeling. The paper also packages the two-tower method, in which a common bi-torsor identifies $N/G'$ with $\\mathrm{Def}^G_{F(X)}$, allowing comparison across different group actions. The Lubin-Tate examples—a supersingular elliptic curve model, a plane-curve model at height $p-1$, and a PEL abelian-variety moduli model at general height $h$—are presented as instances of the same criterion.","pith_inferences":["Editorial: The proof of the criterion is largely formal, so the real mathematical work lies in producing modeling functors; a systematic construction for every maximal finite subgroup of $\\operatorname{Aut}_k(F)$ at every height would turn the framework into a computational tool for the cohomology theories built from formal groups.","Editorial: The two-tower comparison suggests a broader pattern: any pair of commuting torsors over a common space should yield analogous cohomological equivalences, so one could test the method on p-adic period spaces beyond the one used in the paper.","Editorial: The paper leaves open the converse question—whether an equivalence of quotient stacks $\\mathrm{Def}^G_X \\simeq \\mathrm{Def}^G_{F(X)}$ always lifts to a $G$-equivariant star-equivalence; finding a counterexample would draw the boundary of the method."],"forward_implications":["Whenever a modeling functor $F:\\mathcal{M}\\to\\mathcal{M}^{\\natural}_{\\mathrm{fg1}}$ exists, the coherent cohomology of $\\mathrm{Def}^G_X$ is the continuous group cohomology of $G$ on $\\mathcal{O}(\\mathrm{Def}^{\\star}_X)$, so computations reduce to group invariants.","In the formal-group model, the star-deformation ring is $W[[u_1,\\ldots,u_{h-1}]]$ (or with $\\beta^{\\pm 1}$ in the graded case), so the modeled quotient has explicit invariant subring $W[[u_1,\\ldots,u_{h-1}]]^G$ as its global sections.","The two-tower method gives a second route: when a common bi-torsor relates $N$ to $\\mathrm{Def}^{\\star}_X$, the quotient $N/G'$ is equivalent to $\\mathrm{Def}^G_X$, so hard $G$-cohomology can be computed as $G'$-cohomology on $N$.","Since the conclusion is an equivalence of stacks, coherent cohomology with any quasi-coherent sheaf—not just the structure sheaf—is preserved under modeling.","For the Lubin-Tate action, the examples provide working models for supersingular elliptic curves, for height $p-1$, and for general height $h$, showing the framework covers the automorphism group action at several heights rather than one isolated case."],"supporting_citations":[{"why":"Introduces the deformation space of a one-dimensional formal group and its automorphism action, the motivating example of the paper.","marker":"(LT66)"},{"why":"Supplies the identification between deformations of an elliptic curve and deformations of its formal group, the input for the elliptic curve example.","marker":"(CS64)"},{"why":"Supplies the Shimura-variety model for the full automorphism group of a formal group, the key height-$h$ example.","marker":"(Car90)"},{"why":"Supplies the height classification, the ind-etale isomorphism scheme, and the deformation rings used to identify formal group deformations.","marker":"(Pst)"},{"why":"Supplies the graded formal group formalism and representability results for the deformation rings in Section 5.5.","marker":"(Lur)"},{"why":"Supplies the equivalence between smooth representations of a profinite group and sheaves on the classifying stack, used in the cohomological identification.","marker":"(HM)"},{"why":"Supplies the p-adic period spaces used as the bi-torsor input in the two-tower method.","marker":"(RZ96)"}],"fun_headline_variants":["Modeling group actions preserves their cohomology","Swap group actions, keep the cohomology","Lubin-Tate action: find a better model, same cohomology","Cohomology invariant under stack action models","Replace stacks, keep continuous group cohomology"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The entire framework assumes you already have a group-compatible way to replace the object you care about by a one-dimensional formal group without changing its deformation problem; the paper supplies such replacements for elliptic curves, plane curves, and Shimura varieties, but for a new group action finding this replacement is a separate hard problem.","fun_headline_variants_meta":{"raw":{"variants":["Modeling group actions preserves their cohomology","Swap group actions, keep the cohomology","Lubin-Tate action: find a better model, same cohomology","Cohomology invariant under stack action models","Replace stacks, keep continuous group cohomology"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000347,"raw_usage":{"total_tokens":1895,"prompt_tokens":937,"completion_tokens":958,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":553,"completion_tokens_details":{"reasoning_tokens":881}},"tokens_in":553,"tokens_out":958,"duration_ms":10743,"temperature":1.0,"reasoning_tokens":881,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T21:18:43.470242+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete check: take a finite subgroup $G$ of automorphisms of a supersingular elliptic curve $E$ over an algebraically closed field of characteristic $p$, and compare the $G$-invariant subring of the completed local ring of the moduli stack of elliptic curves at $E$ with $W[[u_1]]^G$ for the corresponding height-two formal group. If these invariant subrings are not isomorphic, Corollary 6.1 fails; if they are, the supersingular elliptic curve example is confirmed.","supporting_citations":[],"review_version":1}