{"id":"f19ca8d3-5f74-4cb0-acff-b1580f851778","arxiv_id":"2511.22811","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For p≥7, the neutral components of p-adic monodromy groups of supersingular abelian surfaces over Q_p are classified and shown to be generically isomorphic to GL_2 ×_det GL_2.","lead":"This paper classifies the p-adic filtered vector spaces attached to abelian surfaces over Q_p with supersingular good reduction, for primes p≥7, and computes their algebraic monodromy groups. Generically the connected monodromy group is the fiber product GL_2 ×_det GL_2, with explicit exceptional cases.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Prop 3.15's displayed skew-form construction does not make Fil^1 totally isotropic; the S3 verification underpinning Theorem A is invalid as written.","rationale":"The reader identified Volkov's theorem as the weakest assumption, but the more immediate soft spot is internal: Proposition 3.15 is the step that verifies Volkov's S3 for every admissible S1 module, and its displayed construction fails its own isotropy check. I found an explicit admissible filtered φ-module with Fil^1=span(e0,φ^2e0) where the prescribed equations force δ(e0,φ^2e0)≠0, contradicting the claim that Fil^1 is totally isotropic. This does not by itself disprove the classification—the module may still admit a different skew form—but it means the proof of the geometricity half of Theorem A is not correct as written. Because the central claim depends on that proof, the verdict should be conditional on repairing or replacing the S3 construction. I did not find an independent counterexample to the classification itself; the issue is a concrete gap in a load-bearing proof step.","tokens_in":33034,"tokens_out":54563,"duration_ms":447899,"concrete_test":"For p=7, ε=0, take φ to be the companion matrix of X^4+49 and Fil^1=span(e0,e2). Set α=(0,1,0) and choose x1=1, x2=−7, x3=7 as prescribed by Prop 3.15. Compute the displayed 4×4 matrix δ; verify φ^*δ=7δ and det δ≠0, but δ(e0,e2)=−7≠0, so Fil^1 is not totally isotropic. This settles that the construction in Prop 3.15 fails. Then test the corrected equation (α1+(1−ε)p α3)x1+α2 x2=0; if isotropy holds for all α, the theorem may be repairable, but the printed proof is invalid.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Theorem A's geometricity direction ('every listed module arises from an abelian surface') goes through Volkov's S3 condition, verified in Proposition 3.15. The proof constructs δ using x1,x2,x3 satisfying x3=(1−ε)p x1 and (α1−(ε−1)p α2)x1+α2 x2=0. But for Fil^1=span(x,y) with y≡α1φx+α2φ^2x+α3φ^3x mod x, total isotropy requires δ(x,y)=0, and δ(x,y)=α1x1+α2x2+α3x3. Substituting the displayed equation gives δ(x,y)=(1−ε)p(α3−α2)x1, which is not zero when α2≠α3 and ε≠1. Concretely, take ε=0, φ the companion matrix of X^4+p^2, and Fil^1=span(e0,e2). Then α=(0,1,0), and the system forces p x1+x2=0. Choosing x1=1,x2=−p,x3=p gives a nondegenerate δ with φ as a p-similitude, but δ(e0,e2)=−p≠0, so Fil^1 is not totally isotropic. Thus the proof of S3 as printed is incorrect; the equation likely has α2 in place of α3. Since Theorem A's 'precisely' depends on this verification, the central claim is not fully supported as written.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper classifies, for primes p ≥ 7, the filtered φ-modules attached to the p-adic Tate modules of abelian surfaces over Q_p with supersingular good reduction. The main results are: a complete list of such modules (four explicit families with arithmetical parameter conditions, Theorem A / Theorems 3.7 and 3.10); the determination of the neutral connected components of the associated p-adic algebraic monodromy groups (Theorem B / Theorem 4.6), which are G_m^2, G_m^3, G_a^2 ⋊ G_m^2, GL_2, or GL_2 ×_det GL_2; a moduli-space description of the isomorphism classes via a GIT quotient of Gr(2,4) (Theorem C / 5.7); and a proof that the generic neutral component is GL_2 ×_det GL_2 (Theorem D / 5.13), together with a description of Wintenberger types (Theorem E / 5.8). The proof strategy combines a classification of supersingular p-Weil polynomials of degree 4, linear-algebraic constraints from Hodge decompositions, Volkov's geometricity criterion, Pink's density theorem, and explicit Lie-algebra computations.","tokens_in":33377,"tokens_out":7942,"duration_ms":66345,"significance":"If correct, the paper gives a complete, explicit local classification for a concrete class of geometric Galois representations and computes their monodromy groups, extending Serre's results for elliptic curves to supersingular abelian surfaces. The moduli-space description via a GIT quotient is elegant and makes the distribution of monodromy groups precise; the paper also provides the first examples in this setting where the neutral component is non-reductive. A notable strength is the concreteness of the output: the author gives explicit matrices for all constructed objects, parameter conditions, and Lie-algebra computations, which makes the claims checkable. The dependence on Volkov's external characterization is clearly stated and is a standard tool, not a circularity.","major_comments":[{"comment":"The construction of the skew form does not, as written, make Fil^1 totally isotropic. With y ≡ α1 φx + α2 φ²x + α3 φ³x (mod Q_p x), total isotropy requires δ(x,y)=α1 x1+α2 x2+α3 x3=0. Substituting the displayed conditions x3=(1−ε)p x1 and (α1−(ε−1)p α2)x1+α2 x2=0 gives δ(x,y)=(1−ε)p(α3−α2)x1, which is not zero in general. For instance, take ε=0, φ the companion matrix of X^4+p^2, and Fil^1=span(e0,e2); then α=(0,1,0), and the system forces x2=−p x1, x3=p x1, so δ(e0,e2)=−p x1 ≠ 0. Thus the verification of Volkov's condition S3 is invalid as printed. Since Proposition 3.15 is the step that turns the algebraic classification into geometric objects via Volkov's criterion, Theorem A's 'precisely' and the geometric direction of Theorem 3.10 are not fully supported. The error appears repairable (the correct condition is α1 x1+α2 x2+α3 x3=0, and the example admits the valid choice x2=0, x3=p x1","section":"§3.6, Proposition 3.15"}],"minor_comments":[{"comment":"Typo: 'arithmetric conditions' should be 'arithmetic conditions'.","section":"Abstract and Theorem A"},{"comment":"Typo: 'computating' should be 'computing'.","section":"§2.4 title"},{"comment":"Prop. 3.1 says 'degree 3 smaller than 4'; the wording is confusing, probably 'degree smaller than 4' or 'degree ≤ 4' is intended.","section":"§1.4 and §3.1"},{"comment":"'Exemple' should be 'Example'; 'nonsemisimple' is a typo for 'non-semisimple'.","section":"Example 4.4"},{"comment":"In the case X_D=(A), the deduction that both a and (a−εp)/p lie in pZ_p is terse; a sentence explaining why the image of ar u−Id must be contained in a proper ar f_el-stable subspace would help. This is not load-bearing for the main theorem once Prop. 3.15 is fixed.","section":"§3.5, proof of Proposition 3.14"}],"recommendation":"major_revision","confidential_remarks":"The main issue is exactly the one flagged by the stress-test: the S3 verification in Proposition 3.15 is algebraically wrong as written. I believe the statement is very likely true and the proof can be repaired with a correct choice of x1,x2,x3, but because this verification is the bridge to Volkov's geometricity criterion, the central classification theorem is not yet fully established. I therefore recommend major revision rather than rejection. The rest of the paper is detailed and the monodromy computations are extensive; I did not find other load-bearing errors. The footnote disputing Volkov's Proposition 2.1 is presented rather polemically and may deserve a more neutral formulation, but that is a minor stylistic concern."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is a genuine advance—the first p-adic monodromy classification for a family of abelian varieties beyond elliptic curves—but there is a concrete error in the S3 verification of Proposition 3.15 that, as written, breaks the proof of Theorem A's geometricity direction. It looks like a repairable typo, but needs to be fixed before the main theorem is fully supported.\n\nWhat the paper does well: the classification of supersingular Weil polynomials of degree 4 is clean, and the reduction to linear-algebraic constraints (Lemma 3.2) is sensible. The monodromy computations in Section 4, via Pink's density theorem and Lie algebra generation, are long but systematic; the resulting list (G_m^2, G_m^3, G_a^2 ⋊ G_m^2, GL_2, GL_2 ×_det GL_2) is a satisfying extension of Serre's elliptic curve result. The moduli-space/GIT interpretation in Section 5 is a nice bonus, and the claimed correction of Volkov's earlier claim is worth checking on its own.\n\nWhere the soft spots are: the stress-test note is right. In Proposition 3.15, to prove S3 for the X^4+εpX^2+p^2 family, they choose x1,x2,x3 with x3=(1−ε)p x1 and (α1−(ε−1)p α2)x1+α2 x2=0. But total isotropy of Fil^1 requires δ(x,y)=α1x1+α2x2+α3x3=0, and these equations give δ(x,y)=p(1−ε)(α3−α2)x1, not zero in general. The concrete example in the stress-test (ε=0, companion matrix X^4+p^2, Fil^1=span(e0,e2)) makes this vivid. The likely fix is to put α3 in place of α2 in the first equation; then the conclusion follows. As printed, though, Theorem A's 'precisely' is not fully supported because S3 is not verified. This is small in the sense of being repairable, but it is load-bearing.\n\nOtherwise, the paper is honest about using Volkov's criterion as an external black box, and that is the right way to proceed. The many 'direct calculations' are a bit annoying, but the matrices are mostly given. Typos and formatting issues abound but are minor.\n\nThis paper is for p-adic Hodge theorists and people working on monodromy of abelian varieties. It deserves a serious referee; I would send it out. The referee should ask for a corrected Prop 3.15 and a careful check of the matrix computations in Props 4.2–4.5 and 5.5. With that, this should be a solid accept.","headline":"A strong classification paper with a real but repairable bug in the S3 verification (Prop 3.15) that the referee must catch.","tokens_in":33836,"tokens_out":5108,"would_cite":true,"duration_ms":39922,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["11G10","11F80","14G20","14L24"],"pacs":[],"model":"deepseek-v4-flash","headline":"For primes p ≥ 7, the p-adic monodromy groups of supersingular abelian surfaces over Q_p are classified explicitly: the neutral connected component is one of five algebraic groups, generically GL_2 ×_det GL_2.","keywords":["supersingular abelian surfaces","filtered φ-modules","p-adic monodromy groups","crystalline representations","Tate modules","Wintenberger type","GIT quotient","p-Weil polynomials"],"falsifier":"Exhibit, for some prime p ≥ 7, a filtered φ-module satisfying the three geometric-origin conditions S1–S3 with characteristic polynomial X^4 + ϵpX^2 + p^2 whose isomorphism class is not among the four families, or a supersingular abelian surface whose p-adic monodromy neutral component is not one of the five groups. A computer search over the parameter space (a,b) ∈ Q_p^2 with ab ≠ −1 and the stated valuation conditions could check the predicted monodromy isomorphisms directly.","tokens_in":32932,"feed_emoji":"🧮","tokens_out":5540,"duration_ms":44508,"temperature":0.7,"pith_summary":"This paper aims to give a complete classification of the p-adic linear-algebraic objects (filtered φ-modules) attached to supersingular abelian surfaces over Q_p, for primes p ≥ 7. It shows there are exactly four explicit families of such modules, parametrized by rational parameters satisfying simple valuation conditions. From this, it computes the neutral connected components of the associated algebraic monodromy groups: only five groups occur, and for a generic point in the moduli space the component is GL_2 ×_det GL_2. The result matters because it turns the study of these monodromy groups — which control the image of the Galois action on p-adic Tate modules — into a finite list with an explicit moduli description, a p-adic analogue of Serre's results on elliptic curves.","feed_headline":"Five monodromy groups cover supersingular abelian surfaces","feed_subtitle":"For p≥7, the p-adic Tate modules of supersingular abelian surfaces over Q_p are classified; nearly all have monodromy GL2×_det GL2.","key_machinery":"The central objects are admissible filtered φ-modules: a Q_p-vector space with a Frobenius endomorphism φ and a Hodge filtration, obtained from a p-adic Tate module via Fontaine's functor. The paper's main tools are (1) the rigidity of supersingular p-Weil polynomials of degree 4 for p ≥ 7, which forces the characteristic polynomial of φ to be either (X^2 ± p)^2 or X^4 + ϵpX^2 + p^2; (2) Wintenberger's Hodge decomposition, whose periodic-function labels ('Wintenberger type') control the possible shapes; and (3) a density result showing the monodromy group is generated by conjugates of the Hodge cocharacter by powers of φ. The moduli description uses the Grassmannian Gr(2,4) and a GIT quotien","core_discovery":"The paper establishes that, for p ≥ 7, every filtered φ-module arising from an abelian surface over Q_p with supersingular good reduction is isomorphic to one of four families — a product of two rank-2 modules, an isolated member, and two parameterized families D^ϵ,ν and D^ϵ,μ — with parameters satisfying explicit valuation conditions. For these modules, the neutral connected component of the monodromy group is, over the algebraic closure, one of G_m^2, G_m^3, G_a^2 ⋊ G_m^2, GL_2, or GL_2 ×_det GL_2. Moreover, in the coarse moduli space of such modules, all but finitely many points have the generic component GL_2 ×_det GL_2; the exceptional loci are described precisely, and the Wintenberger","pith_inferences":["One testable extension is to small primes: the arguments exclude p < 7 at specific steps (e.g., roots of X^2 + ϵpX + 1), so the five-group list may change for p = 3, 5; computing the monodromy groups there would show how the classification bifurcates.","The generic-large-monodromy statement suggests a p-adic analogue of Serre's open image theorem for abelian surfaces: the exceptional locus in the moduli space is finite, so deformations of a supersingular surface should almost always have full monodromy.","The moduli picture may transfer to the study of Hecke orbits: since monodromy groups are stable under Hecke correspondences, the finite-exception statement gives local evidence for p-adic nowhere-density of small monodromy loci in Hodge-type Shimura varieties.","The method of classifying via Wintenberger types could be pushed to higher-dimensional abelian varieties with supersingular reduction; the first obstruction would be the larger set of supersingular Weil polynomials, but the monodromy computation via conjugates of the Hodge cocharacter should remain tractable."],"forward_implications":["Every supersingular abelian surface over Q_p (p ≥ 7) has p-adic monodromy neutral component in the five-group list, so the image of the Galois representation is understood up to finite index and inner forms.","The generic component GL_2 ×_det GL_2 is as large as the Hodge and determinant constraints allow; non-generic behaviour is confined to finitely many loci in the moduli space.","Non-semisimple p-adic Tate modules occur only when p ≡ 1 mod 3 and the module is the specific member D^{1,μ}_{(−ζ_3 p,1)}; correspondingly, non-reductive monodromy occurs only for the G_a^2 ⋊ G_m^2 case.","The moduli space M^wa is a quotient of P^2(Q_p) by the centralizer of φ, with a projective GIT quotient map to P^1(Q_p); the Wintenberger type is constant above the same valuation of a single function.","The classification of supersingular p-Weil polynomials of degree 4 for p ≥ 7 gives only the two families (X^2 ± p)^2 and X^4 + ϵpX^2 + p^2, a purely arithmetic fact that underpins the whole paper."],"fun_headline_variants":["Supersingular surfaces: five monodromy types, generic GL2×GL2","Supersingular abelian surfaces: monodromy generically GL2×GL2","Generic monodromy for supersingular abelian surfaces is GL2×GL2","Classification reveals generic GL2×GL2 monodromy for supersingular surfaces"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The classification of which filtered φ-modules actually come from abelian surfaces with good reduction is taken from an imported theorem in the literature (Conditions S1–S3); if that geometric-origin criterion is incomplete, the 'precisely' in the classification and the claim that the listed modules all arise from abelian surfaces would lose their footing.","fun_headline_variants_meta":{"raw":{"variants":["Supersingular surfaces: five monodromy types, generic GL2×GL2","Supersingular abelian surfaces: monodromy generically GL2×GL2","Generic monodromy for supersingular abelian surfaces is GL2×GL2","Classification reveals generic GL2×GL2 monodromy for supersingular surfaces"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001365,"raw_usage":{"total_tokens":5354,"prompt_tokens":708,"completion_tokens":4646,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":452,"completion_tokens_details":{"reasoning_tokens":4550}},"tokens_in":452,"tokens_out":4646,"duration_ms":30075,"temperature":1.0,"reasoning_tokens":4550,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T19:40:25.224752+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Exhibit, for some prime p ≥ 7, a filtered φ-module satisfying the three geometric-origin conditions S1–S3 with characteristic polynomial X^4 + ϵpX^2 + p^2 whose isomorphism class is not among the four families, or a supersingular abelian surface whose p-adic monodromy neutral component is not one of the five groups. A computer search over the parameter space (a,b) ∈ Q_p^2 with ab ≠ −1 and the stated valuation conditions could check the predicted monodromy isomorphisms directly.","supporting_citations":[],"review_version":1}