{"id":"1d05826a-b8eb-4db3-8dc9-56155f7b7c88","arxiv_id":"2509.01573","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Over arbitrary p-adic formal bases, finite locally free p-power torsion commutative group schemes are exactly the perfect F-gauges of Tor amplitude [-1,0] and Hodge-Tate weights 0,1, via an exact, Cartier-duality-compatible equivalence.","lead":"Finite flat group schemes over p-adic rings are shown to be classified exactly by certain cohomological objects called perfect F-gauges, for any p-adic base. This gives one canonical framework that unifies and extends prior classifications by Kisin, Breuil, Zink, Lau, and Anschütz-LeBras, with new consequences for cohomology and purity.","discovery_kind":"unification","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem A inherits its force from the imported [23, Thm 11.1.4] and the qrsp-exactness import [43]; any unstated restriction there breaks the five-lemma in §7.3 and the exactness reduction in §7.4.","rationale":"The reader's CONDITIONAL verdict already identifies the same load-bearing assumption: the imported Theorem 4.3.1 ([23, Thm 11.1.4]), together with the representability results and the qrsp exactness import from [43]. My stress-test agrees and adds that the topology of Raynaud's theorem used in §7.3 also needs an explicit descent argument if it is only fppf. The proof of Theorem A is otherwise internally coherent: the functor G is canonical, Theorem 6.2.1 supplies the needed F-gauge analogue of Raynaud, and the exactness reductions are carefully structured. The risk is external rather than internal, and it is exactly the kind of risk that justifies the reader's CONDITIONAL verdict rather than ACCEPT or REJECT.","tokens_in":69358,"tokens_out":26746,"duration_ms":323677,"concrete_test":"Audit [23, Theorem 11.1.4] and its proof: list every base hypothesis used and verify that it covers all p-complete discrete rings, including a non-qrsp example such as R = Z_p[x]^\\wedge, and that the stack isomorphism Vectsyn_{n,{0,1}} ≅ BT_n is proved over Z_p without a qrsp-only reduction. Then re-run the five-lemma of Lemma 7.3.1 with M1 = cofib(V^{-1}→V^0) and M2 = W a vector bundle F-gauge, using only Theorem 4.3.1 and Theorem 5.1.1; if either theorem has an unstated restriction that fails for Z_p[x]^\\wedge, Theorem 7.1.1's full faithfulness and hence the main equivalence fail. For the Raynaud step, check the topology of the cover in [6, Théorème 3.1.1]; if it is fppf, write the descent datum on the F-gauge explicitly and verify effectiveness via fpqc descent for P_syn_{0,1}. If no such descent is possible, essential surjectivity is not established.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The decisive steps in §7.3 and §7.4 are reductions to prior results, not new arguments. Lemma 7.3.1 proves full faithfulness by a five-lemma in which the outer vertical maps are isomorphisms by Theorem 4.3.1 ([23, Thm 11.1.4]) and Theorem 5.1.1. Theorem 5.1.1 itself relies, via Remark 5.1.3, on the qrsp case imported from [43], and then on Theorem 5.2.2. Thus if [23, Thm 11.1.4] is not valid for every p-complete discrete ring as stated, or if its compatibility with arbitrary base change was only proved under hidden hypotheses, the five-lemma in Lemma 7.3.1 collapses, and with it essential surjectivity and exactness in Theorem 7.1.1. A second, closely related pressure point is essential surjectivity's use of Raynaud's theorem [6, Théorème 3.1.1]: the text says the group scheme may be assumed étale-locally a kernel of an isogeny of p-divisible groups, but the classical theorem is usually fppf-local. If the cover is only fppf, the proof must explicitly descend the F-gauge constructed over the cover; this descent step is not written. Both issues are about the external platform, not an internal inconsistency in the checkable portions of the paper.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves an exact, canonical, base-change-compatible equivalence of categories G : P_syn_{0,1}(R) -> FFG(R) for p-complete discrete rings R, carrying p^n-torsion F-gauges onto p^n-torsion finite flat commutative group schemes and compatible with Cartier duality (Theorem 7.1.1). The proof route is: (i) define G via truncated syntomic cohomology; (ii) prove exactness of Gamma_syn for truncated Barsotti-Tate groups (Theorem 5.1.1); (iii) prove an F-gauge analogue of Raynaud's theorem (Theorem 6.2.1); (iv) assemble full faithfulness, essential surjectivity, and exactness of Theorem 7.1.1. The paper also proves a syntomic/fppf cohomology comparison, representability of relative fppf cohomology along proper smooth maps, a Česnavičius-Scholze purity theorem, and several explicit classifications in terms of divided Dieudonné complexes and Breuil-Kisin frames.","tokens_in":69627,"tokens_out":14070,"duration_ms":161344,"significance":"If the proof is completed as intended, this is a substantial unification: finite flat group schemes over arbitrary p-adic bases are classified by one canonical cohomological object, specializing to truncated Barsotti-Tate groups, qrsp/prismatic Dieudonné theory, windows, and Breuil-Kisin modules. The paper is well structured and I did not find internal circularity: Theorems 5.1.1 and 6.2.1 are proved before and independently of Theorem A, and Theorem A is assembled from them by explicit five-lemma and descent arguments. The exactness theorem and the F-gauge Raynaud theorem are themselves valuable. The main risks are external dependencies and one or two unwritten descent steps; these are fixable within the manuscript's scope.","major_comments":[{"comment":"The text says, after proving full faithfulness, that by Raynaud's theorem [6, Théorème 3.1.1] one may assume the group scheme is étale-locally a kernel of an isogeny of p-divisible groups. As stated in [6], the classical theorem gives an fppf-local presentation, not an étale-local one. If the cover is only fppf, one must descend the F-gauge V and the isogeny data along that cover using fpqc/fppf descent. At this point in the paper fpqc descent for P_syn_{0,1} has not yet been established, and the descent step is not written. Since this is the final step of essential surjectivity, add an explicit descent argument or give a reference for an étale-local version.","section":"§7.3, essential surjectivity"},{"comment":"Theorem 5.1.1 is load-bearing: it supplies the outer vertical isomorphisms in the five-lemma of Lemma 7.3.1 and the final exactness assertion in §7.4. Its proof reduces to the qrsp case by invoking [43, Remark 3.83] and [43, §3]. The exact content of the imported statement—in particular, that the inverse functor carries short exact sequences of finite flat group schemes to fiber sequences of perfect F-gauges for every qrsp ring, compatibly with Gamma_syn—is not restated. A hidden hypothesis there would propagate directly to Theorem A. Please state the imported theorem precisely, or prove the needed qrsp case. Also make explicit the step from isomorphism on qrsp R-algebras to an isomorphism of formal stacks in (5.1.4.1).","section":"§5.1 / Remark 5.1.3"},{"comment":"Theorem 8.2.1 depends on the bound that Rπ^syn_* M is again a perfect F-gauge with Hodge-Tate weights in [n-d,m] and Tor amplitude [a,b+2d]. The proof is only sketched ('can be deduced from results of Guo-Li'), and the final reduction to coherent cohomology of proper morphisms is indicated rather than proved. Since Theorem C is advertised as a main application, please provide a complete proof or a precise reference that covers F-gauges over p-adic formal algebraic spaces in the required generality.","section":"§8.2 / Proposition 8.2.6"}],"minor_comments":[{"comment":"The introduction says Raynaud's theorem gives a Zariski-local reduction, while §7.3 says étale-locally. Align these statements once the correct topology is fixed.","section":"§1.4 vs §7.3"},{"comment":"The notation P^syn_{[0,1],n}(R) appears to conflict with the paper's usual P^syn_{n,{0,1}}(R). Please clarify which category is meant in this lemma.","section":"Lemma 6.3.3"},{"comment":"The body states Theorem A for affine Spf R, while the introduction states it for arbitrary p-adic formal schemes. The globalization by Zariski/fpqc descent is asserted rather than demonstrated; a short explanation would help.","section":"Theorem 7.1.1 / Remark 1.1.3"},{"comment":"The identification Gamma_syn(O_syn{1}[1]) ≃ lim_{←m} B μ_{p^m} is used for Cartier duality. Please give a reference or a proof for this identification.","section":"Remark 7.2.2"},{"comment":"The reduction from a general qcqs scheme to 'X lci of dimension ≤ d' via [15, Lemma 7.1.1] is too terse; please spell out the constructibility and dimension hypotheses.","section":"Corollary 8.3.7"},{"comment":"The statement 'R^∆ is already classical [11, Corollary 8.13]' should be made precise: classical as a prestack, as a ring object, or as having discrete structure sheaf on semiperfectoid test objects? This is needed for the advertised reduction to classical divided Dieudonné complexes.","section":"§9.9.7(1)"}],"recommendation":"major_revision","confidential_remarks":"The main risk to Theorem A is the external platform: [23, Thm 11.1.4], the qrsp exactness statement from [43], and Raynaud's theorem. I did not find internal circularity, but the paper should be asked to state the imported results in the exact generality used and to fill the fppf-to-étale descent gap in §7.3. If those are supplied, the central claim is very likely sound."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my read. Theorem A genuinely does what it says: it extends the F-gauge classification of finite flat group schemes from truncated BT groups and quasisyntomic bases to arbitrary p-complete discrete rings, and the exactness theorems (5.1.1, 5.2.2) plus Raynaud-for-F-gauges (6.2.1) are substantive results, not packaging. The authors are honest about provenance: they flag what comes from [23], what comes from [43], and which §9 results are recoveries. The proof structure is non-circular in the checkable parts, and the cohomological applications in §8 look like real payoff.\n\nThe soft spots are all upstream dependencies. Full faithfulness and essential surjectivity in §7.3 close via the imported Theorem 4.3.1 ([23, Thm 11.1.4]) and Theorem 5.1.1, which itself imports the qrsp case from [43] (Remark 5.1.3). So the five-lemma in Lemma 7.3.1 inherits any hidden hypotheses in those statements. I don't see evidence of such hypotheses, but a referee should explicitly verify that [23]'s equivalence is compatible with arbitrary base change in exactly the stated generality, because the proof as written does not re-prove it.\n\nThe second point is the one I most want the referee to press. Essential surjectivity invokes classical Raynaud's theorem [6, Théorème 3.1.1] and says the group scheme can be assumed to be a kernel of an isogeny of p-divisible groups. As usually stated, Raynaud gives an fppf cover, not a Zariski one. If the cover is only fppf, the F-gauge constructed over the cover must be descended, and that descent step is not written. It is probably fillable given the fpqc descent that Theorem A itself claims, but it needs to be in the paper.\n\nSmaller issues: Proposition 8.2.6, which carries Theorem C, is a sketch with an explicit reliance on Guo–Li results \"not stated in the generality we require\"; Remark 7.1.2 announces an analogue of Prop 6.4.3 without proof; and the final subsection 9.11 was truncated in the version I saw, so the last claimed classifications could not be inspected. These are blemishes, not sinkholes.\n\nOverall: the architecture is coherent, the central theorem is probably correct, and the paper deserves a serious referee. I would send it out, and I would want the referee to pin down the Raynaud descent point and the exact hypotheses imported from [23].","headline":"A serious, likely-correct extension of the Bhatt-Lurie/Anschütz-Le Bras classification that deserves refereeing, but the decisive reductions lean heavily on imported [23] and [43] and a couple of descent/approximation steps are left unwritten.","tokens_in":70411,"tokens_out":3202,"would_cite":true,"duration_ms":39694,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14L15","14F30","11G10"],"pacs":[],"model":"deepseek-v4-flash","headline":"Perfect F-gauges classify finite flat p-group schemes","keywords":["finite flat group schemes","perfect F-gauges","syntomification","prismatic cohomology","Dieudonné theory","Breuil-Kisin modules","p-divisible groups","fppf cohomology"],"falsifier":"Take a non-perfect p-adic ring such as F_p[[t]] and compare Hom between two explicit F-gauges M_1, M_2 with Hom between the associated group schemes G(M_1), G(M_2), computed through τ≤0RΓ(R_syn, M); if a nonzero F-gauge map became zero at the level of group schemes, full faithfulness would fail. For exactness, compare the extension group Ext(G(M_1),G(M_2)) in fppf groups with Hom(M_1[1],M_2) in perfect F-gauges over a base where both can be computed by matrices, such as a semiperfectoid ring with non-F-nilpotent reduction; a mismatch would disprove Theorem A. A simpler global check is whether","tokens_in":69004,"feed_emoji":"","tokens_out":9121,"duration_ms":103087,"temperature":0.7,"pith_summary":"This paper proves that finite locally free commutative p-power torsion group schemes over any p-complete discrete base ring are classified, by a canonical exact equivalence, by perfect F-gauges: perfect complexes on the syntomification of the base with Hodge-Tate weights in {0,1}, Tor amplitude in [-1,0], and p-power-torsion cohomology. The equivalence is compatible with arbitrary base change, fpqc descent, and Cartier duality, and it turns fppf cohomology of the group scheme into syntomic cohomology of the F-gauge. If correct, this gives one classification that specializes to all known frameworks such as windows, displays, crystals, and prismatic φ-modules, and it extends them past truncated Barsotti-Tate groups to general bases. It also yields representability of relative fppf cohomology under proper smooth maps and a proof of a purity theorem for fppf cohomology.","feed_headline":"Perfect F-gauges classify finite flat p-group schemes","feed_subtitle":"A single canonical functor over p-adic bases unifies every known classification of these group schemes.","key_machinery":"The central object is a perfect F-gauge: a dualizable quasicoherent sheaf on the syntomification R_syn, the p-adic cohomological stack whose coherent cohomology computes syntomic cohomology, equipped with Hodge-Tate weights 0,1, Tor amplitude [-1,0], and p-power-torsion cohomology. The identity doing the work is the cofiber presentation: pro-étale locally, every such M is cofib(V^{-1} → V^0) with V^{-1}, V^0 vector-bundle F-gauges, and under the prior vector-bundle classification these correspond to an isogeny H^{-1}→H^0 of p-divisible groups, so G(M) is realized as the finite flat kernel of that isogeny. The marked twists M∨{1}[1] and the duality pairing G(M∨{1}[1]) ≃ G(M)^* are the Cartier","core_discovery":"The central claim is Theorem A: over every p-complete discrete ring R, the category FFG(R) of finite locally free p-power-torsion commutative group schemes is equivalent to the category P_syn_{0,1}(R) of perfect F-gauges over R_syn with Hodge-Tate weights in {0,1}, Tor amplitude in [-1,0], and cohomology killed by a power of p. The functor G is truncated syntomic cohomology: on p-nilpotent test rings C, G(M)(C) = τ≤0 RΓ(C_syn, M|C_syn). The proof runs by combining an F-gauge analogue of Raynaud's theorem—every perfect F-gauge of the right weights is pro-étale locally the cofiber of a map V^{-1}→V^0 of vector-bundle F-gauges—with the known classification of n-truncated Barsotti-Tate groups by","pith_inferences":["If the canonical equivalence is as functorial as stated, the inverse functor may be definable on all p-adic formal stacks through moduli of classifying stacks and Nygaard-filtered prismatic cohomology, not only on p-quasisyntomic bases; this would turn the classification into a computational tool for finite flat group schemes in families.","Since the equivalence satisfies fpqc descent and is compatible with arbitrary base change, it likely globalizes to p-adic formal algebraic stacks and can transfer representability statements between syntomic cohomology and fppf cohomology beyond proper smooth morphisms.","The condition that R/pR be F-finite and F-nilpotent is probably sufficient rather than necessary for the classical-truncation classification; testing whether local nilpotence of the divided Frobenius on the cotangent complex alone yields the equivalence would be a direct extension of Theorem E."],"forward_implications":["Every finite flat p-power-torsion group scheme over any p-complete base has an associated perfect F-gauge, and its fppf cohomology is canonically the syntomic cohomology of that gauge.","The category of such group schemes becomes an exact category whose extensions are fiber sequences of F-gauges, so exactness of the equivalence means extension and obstruction problems can be computed in perfect complexes.","All existing classifications—windows, displays, divided crystals, and prismatic φ-modules—are special cases of one canonical functor, and new classifications follow in cases not previously covered.","Relative fppf cohomology of finite flat group schemes under proper smooth maps is representable once lower direct images are, generalizing previously known field and height-one cases.","The purity theorem for fppf cohomology follows by comparing fppf cohomology with syntomic cohomology of the associated F-gauge."],"supporting_citations":[{"why":"supplies the classification of n-truncated Barsotti-Tate groups by vector-bundle F-gauges (Theorem 4.3.1) and the representability theorems that the proof reduces to.","marker":"[23]"},{"why":"provides the explicit inverse of G over quasisyntomic rings and exactness for quasiregular semiperfectoid rings, the base case used in the reduction.","marker":"[43]"},{"why":"gives Raynaud's theorem that every finite flat p-power-torsion group scheme is Zariski locally the kernel of an isogeny of p-divisible groups, used for essential surjectivity.","marker":"[6]"},{"why":"constructs the syntomification and prismatic cohomology underlying F-gauges, and its argument is adapted to prove the purity corollary.","marker":"[10]"},{"why":"supplies the prismatization and Cartier-Witt divisor formalism in which F-gauges and frames are defined.","marker":"[11]"},{"why":"provides divided Dieudonné crystals and F-finite/F-nilpotent constructions used to prove the classical-prismatization classification (Theorem E).","marker":"[33]"},{"why":"gives the existing prismatic Dieudonné classification over qrsp rings that Theorem E generalizes and with which the result is compared.","marker":"[2]"},{"why":"is the purity theorem for fppf cohomology that the paper recovers via the syntomic-fppf comparison.","marker":"[15]"}],"fun_headline_variants":["Perfect F-gauges equal finite flat p-group schemes","New p-adic equivalence: F-gauges and flat group schemes","Finite flat group schemes classified by F-gauges","A categorical bridge: perfect F-gauges and p-torsion groups","F-gauges unify finite flat group scheme classifications"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The load-bearing premise is the imported equivalence, from the paper's prior work, between n-truncated Barsotti-Tate groups and vector-bundle F-gauges over p-complete rings (with the representability theorems behind it); the new proof's full faithfulness, exactness, and essential surjectivity reduce at the final step to this equivalence, and the essential-surjectivity argument also relies on Raynaud's theorem that every finite flat group scheme is locally a kernel of an isoge","fun_headline_variants_meta":{"raw":{"variants":["Perfect F-gauges equal finite flat p-group schemes","New p-adic equivalence: F-gauges and flat group schemes","Finite flat group schemes classified by F-gauges","A categorical bridge: perfect F-gauges and p-torsion groups","F-gauges unify finite flat group scheme classifications"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000856,"raw_usage":{"total_tokens":3564,"prompt_tokens":762,"completion_tokens":2802,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":506,"completion_tokens_details":{"reasoning_tokens":2709}},"tokens_in":506,"tokens_out":2802,"duration_ms":20938,"temperature":1.0,"reasoning_tokens":2709,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T12:25:34.025983+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a non-perfect p-adic ring such as F_p[[t]] and compare Hom between two explicit F-gauges M_1, M_2 with Hom between the associated group schemes G(M_1), G(M_2), computed through τ≤0RΓ(R_syn, M); if a nonzero F-gauge map became zero at the level of group schemes, full faithfulness would fail. For exactness, compare the extension group Ext(G(M_1),G(M_2)) in fppf groups with Hom(M_1[1],M_2) in perfect F-gauges over a base where both can be computed by matrices, such as a semiperfectoid ring with non-F-nilpotent reduction; a mismatch would disprove Theorem A. A simpler global check is whether","supporting_citations":[{"cited_title":"DieudonnétheoryviacohomologyofclassifyingstacksII","cited_arxiv_id":null,"evidence_quote":"provides the explicit inverse of G over quasisyntomic rings and exactness for quasiregular semiperfectoid rings, the base case used in the reduction."},{"cited_title":"Théorie de Dieudonné cristalline","cited_arxiv_id":null,"evidence_quote":"gives Raynaud's theorem that every finite flat p-power-torsion group scheme is Zariski locally the kernel of an isogeny of p-divisible groups, used for essential surjectivity."},{"cited_title":"Prismatic Dieudonné theory","cited_arxiv_id":null,"evidence_quote":"gives the existing prismatic Dieudonné classification over qrsp rings that Theorem E generalizes and with which the result is compared."},{"cited_title":"Purity for flat cohomology","cited_arxiv_id":null,"evidence_quote":"is the purity theorem for fppf cohomology that the paper recovers via the syntomic-fppf comparison."}],"review_version":1}