{"id":"3bafde25-5ea3-4e50-a149-e4973e79ea5f","arxiv_id":"2607.08660","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"Crystalline Galois representations acquire mu_p-equivariant Breuil-Kisin module reductions, yielding the ↑ constraint on inertial weights and proving weight elimination for a general Serre weight conjecture.","lead":"This paper proves new constraints on how crystalline p-adic Galois representations reduce modulo p, using a hidden symmetry (mu_p-equivariance) of Breuil-Kisin modules. The result gives a uniform weight elimination theorem for Serre's conjecture across all unramified reductive groups.","discovery_kind":"unclear","skeptic_critique":{"model":"glm-5.2","headline":"No significant objection identified. The key step (Theorem 4.2.3) has two independent proofs; the reader's concern about Theorem 4.5.12 is mitigated by the prismatization proof in §4.4.","rationale":"The paper presents a well-structured argument with two independent proofs of the critical step (Theorem 4.2.3). The reader's concern about Theorem 4.5.12 is real but not load-bearing, because the prismatization proof in §4.4 provides a separate route that does not depend on specializability. I verified the key computations in Proposition 4.4.1 (μ_p-invariance, Frobenius compatibility, mod p agreement, flatness) and they check out. The combinatorial machinery in §2 is built on established results (Riche-Williamson, Wang). The L-group adaptation in §5 correctly handles γ-twisted conjugation. The explicit GL_2 example confirms consistency with known results. The main residual risk is in the stack-theoretic constructions (particularly the divisorial description of Z^N_p in Proposition 4.6.2 and the map f: Y → W(k)^N in Example 4.6.13), but these are used only in the syntomification proof, not the prismatization proof. Confidence should remain MODERATE — the arguments are sophisticated and full verification requires deep expertise in prismatic cohomology and the prismatization — but no specific defect was identified.","tokens_in":72512,"tokens_out":6778,"duration_ms":417762,"concrete_test":"Verify Proposition 4.4.1 in a non-split case where L ≠ Q_p (e.g., bG = GL_2 × GL_2 with L = Q_{p^2} and γ swapping factors, as in Remark 5.5.11). Specifically, check that the map ρ† constructed from [u]-zV(1) still satisfies the Frobenius compatibility φ∘ρ† = ρ_std∘q when the γ-action on bG is non-trivial, and that the mod p agreement (part 3) holds after restriction of scalars as in Theorem 5.6.2. If the compatibility fails in this case, the restriction-of-scalars extension of the main theorem would be in jeopardy.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The reader identifies Theorem 4.5.12 (specializability via syntomification) as the load-bearing step, but this overstates the risk: Theorem 4.2.3 has two independent proofs. The prismatization proof (§4.4) does not use Theorem 4.5.12 at all. It constructs the map ρ†: Spf(S)/μ_p → W(k)^∆ via the element [u]-zV(1) ∈ W(S), and verifies: (a) μ_p-invariance (using F([ζ])=1 for ζ∈μ_p), (b) Frobenius compatibility φ∘ρ† = ρ_std∘q (via F([u]-zV(1)) = w(φ(E))), (c) mod p agreement with ρ_std, and (d) faithful flatness. The generic relative position follows because both ρ_std and ρ† are flat covers of W(k)^∆, so the relative position of the pulled-back isogeny is intrinsic (Example 4.4.3). The syntomification proof (§4.8) provides a stronger, intrinsic characterization of N but is not on the critical path for the main theorem. The remaining steps — the combinatorial ↑ constraint from μ_p-fixed Grassmannian geometry (§2, building on [RW22], [Wan23]) and the Chen-Nie adaptation to L-groups (§5.5, handling γ-twisted Frobenius conjugation) — are carefully executed and rest on established results. The explicit GL_2 check in the introduction (four inertial weight possibilities matching Berger-Breuil) provides concrete confirmation. No internal inconsistency or hidden circularity was found.","agreement_with_reader":"partial"},"referee_report":{"model":"glm-5.2","summary":"This paper establishes new constraints on the inertial weights of the mod p reduction of a crystalline p-adic Galois representation with prescribed Hodge-Tate weights. The main result (Theorem 5.5.6) states that if ρ: Gal_{Q_p} → ^LG(Z_p) is a crystalline L-parameter with Hodge-Tate weights μ(ρ), then there exist λ ∈ X*(^bT) with λ^dom ↑ μ(ρ) and w ∈ W such that ρ^ss|_{I_{Q_p}} ≅ τ(λ,w), where ↑ is the partial order from the representation theory of reductive groups in characteristic p. The proof proceeds by showing that the mod p reductions of Breuil-Kisin modules attached to crystalline representations acquire a natural μ_p-equivariant structure (Theorem 4.2.3), and then combining this with a geometric study of μ_p-fixed points in affine Grassmannians (§2) to derive the ↑ constraint. The paper also formulates an explicit Serre weight conjecture (Conjecture 6.3.4) for unramified reductive groups and proves the elimination direction. Two independent proofs of the key structural theorem are provided: one via the prismatization (§4.4) and one via the syntomification (§4.8).","tokens_in":72969,"tokens_out":1288,"duration_ms":253375,"significance":"This is a substantial contribution to the interface of p-adic Hodge theory and the mod p representation theory of reductive groups. The key innovation — that crystalline Breuil-Kisin modules acquire a μ_p-equivariant structure — is a genuinely new structural insight, and its geometric origin (the map ρ†: Spf(S)/μ_p → W(k)^∆ in Proposition 4.4.1) is both natural and surprising. The paper provides two independent proofs of the central structural result, which significantly strengthens confidence in its correctness: the prismatization proof (§4.4) is short and direct, while the syntomification proof (§4.8) provides a more intrinsic characterization. The combinatorial core (§2) is developed carefully and self-containedly, building on results of Riche-Williamson [RW22] and Wang [Wan23]. The application to Serre weight conjectures (§6) and the comparison with the Gee-Herzig-Savitt predictions (Proposition 6.4.6) give the results concrete arithmetic significance. The explicit GL_2 verification matching the Berger-Breuil classification provides a welcome sanity check. The generality of the formulation (L-parameters for unramified reductive groups) and the absence of genericity assumptions,","major_comments":[{"comment":"§4.8, end of proof of Theorem 4.5.12: The key compatibility r*_K((id_R)_*E_N)^alg ≅ j*(f*E_N)^alg is established by tracing through a large commutative diagram. While the individual steps are indicated, the final verification that the two paths through the diagram agree is stated rather than spelled out. Given that this is the load-bearing step for the syntomification proof, a sentence or two making the final identification explicit would strengthen the argument. This is not a correctness concern — the prismatization proof in §4.4 provides an independent route that does not use Theorem 4.5.12 — but would improve the readability of the more conceptual proof.","section":null}],"minor_comments":[{"comment":"§1.2, Theorem 1.2: The notation τ(λ,w) is used before its definition in §5.3.9. A forward reference would help the reader.","section":null},{"comment":"§2.3, Figure 2.3.4: The figure is informative but the labeling of alcoves could be clearer; some labels overlap in the rendering.","section":null},{"comment":"§4.6.6, item (3): The construction of the isomorphism Z^∆_p ≃ (Z^N_p)_{y_univ≠0} in both directions is described in detail, but the overall logical flow would benefit from a brief summary sentence stating which direction is used where in the sequel.","section":null},{"comment":"§5.5, equation (5.5.5): The decomposition M ∼ u^{-μ} j u_μ · u^{θ-μ+pwγμ} t w is central to the proof of Proposition 5.5.4. The role of the Iwahori element y and the application of Lemma 5.5.3 could be flagged more prominently, as they are the key inputs from Chen-Nie [CN22].","section":null},{"comment":"§6.3, Conjecture 6.3.4: The conjecture W(ρ) = W_cris(ρ) = W_explicit(ρ) is stated for semisimple ρ. It would be useful to briefly comment on whether the authors expect the equality W_cris = W_explicit to hold without the p-restricted hypothesis, or whether this is genuinely a restricted-range phenomenon.","section":null},{"comment":"Typo in §4.1: 'the Breuil–Kisin prism and its µp-action' — the section title uses µp but the body text sometimes writes μ_p; consistency would be welcome.","section":null},{"comment":"§4.10.13: The proof of Proposition 4.10.13 was found 'with assistance from ChatGPT-5.4 Pro' (acknowledged in §1.8). The proposition is a purely combinatorial lemma with a self-contained proof; the acknowledgment is appropriate and transparent.","section":null}],"recommendation":"minor_revision","confidential_remarks":"The paper is well-suited for a top journal. The two independent proofs of the central theorem are a significant strength. The reader's concern about Theorem 4.5.12 as a load-bearing step is partially mitigated by the prismatization proof, but the syntomification proof is the more conceptual one and deserves to be fully rigorous. I recommend minor revision primarily to address the presentation of the end of the proof in §4.8 and the minor points listed above. No deep structural concerns were identified."},"author_rebuttal":null,"desk_editor":{"model":"glm-5.2","letter":"The headline result is Theorem 5.5.6: for a crystalline L-parameter ρ with Hodge–Tate weights μ(ρ), the semisimple mod p restriction to inertia satisfies ρ^ss|_{I_{Q_p}} ≅ τ(λ, w) for some λ with λ^dom ↑ μ(ρ). This gives a uniform weight elimination result for Serre's conjecture across all unramified reductive groups, replacing prior case-by-case approaches. The ↑ constraint comes from geometry of μ_p-fixed points in affine Grassmannians, not from fitting inertial weights to data, so there's no circularity concern. The GL_2 sanity check in the introduction — recovering exactly the four Berger–Breuil possibilities — is a nice concrete confirmation that the machinery produces the right answer in known cases. The combinatorial core (Section 2) is carefully done, building on Riche–Williamson and Wang's results with self-contained proofs of the key orbit-closure comparison (Proposition 2.4.10). The Chen–Nie adaptation to L-groups in Section 5.5, handling γ-twisted Frobenius conjugation, is carried out cleanly. The reader flags Theorem 4.5.12 (specializability via syntomification) as the load-bearing step. I think this overstates the risk. Theorem 4.2.3 — the actual structural result needed downstream — has two independent proofs. The prismatization proof in §4.4 does not use Theorem 4.5.12 at all; it constructs the map ρ† directly via the element [u]−zV(1) ∈ W(S) and verifies μ_p-equivariance, Frobenius compatibility, mod p agreement with ρ_std, and faithful flatness by hand. The syntomification proof in §4.8 gives a more intrinsic characterization of the modification N but is supplementary, not on the critical path. That said, the stack-theoretic arguments in both §4.4 and §4.8 are dense. The prismatization proof is short and the key computations are explicit, but a referee with expertise in prismatic geometry should verify Proposition 4.4.1 in detail — particularly the faithful flatness argument, which reduces mod p and then lifts via the local flatness criterion. The divisorial description of Z^N_p in §4.6 is a useful reformulation but adds length. This is a strong paper proving a genuinely new result with a new technique. It deserves a serious referee who can check the prismatic geometry carefully. The overall logical structure is sound: two proofs of the key step, established combinatorial foundations, and a concrete known-case check.","headline":"New structural result on crystalline representation reductions, with two independent proofs of the key step","tokens_in":73440,"tokens_out":622,"would_cite":true,"duration_ms":209478,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"glm-5.2","headline":"μ_p-symmetry constrains how crystalline Galois representations reduce mod p","keywords":[],"falsifier":"If there exists a crystalline representation whose Breuil–Kisin module's canonical modification (from the Hodge filtration) fails to be congruent to the Frobenius modification modulo p, then the μ_p-equivariant structure on M/pM is not established, the ↑ constraint on inertial weights need not hold, and the weight elimination result would fail for that representation.","tokens_in":72740,"feed_emoji":"🔒","tokens_out":1096,"duration_ms":268443,"temperature":0.7,"pith_summary":"When a crystalline Galois representation of Q_p — one coming from p-adic Hodge theory with well-defined Hodge–Tate weights — is reduced modulo p, the resulting mod p representation carries only partial memory of those weights. This paper identifies a new structural property of the reduction: the Breuil–Kisin module attached to a crystalline representation acquires a natural equivariance under the group scheme μ_p of p-th roots of unity when reduced modulo p. This μ_p-equivariance, combined with the geometry of μ_p-fixed points in affine Grassmannians, forces the inertial weights of the mod p reduction to satisfy a partial order constraint called the ↑ relation. Specifically, if ρ is crystalline with Hodge–Tate weights μ, then the restriction of the semisimplified mod p reduction to inertia is of the form τ(λ, w) where λ ↑ μ in the p-dilated affine Weyl group. This is established for L-parameters of arbitrary unramified connected reductive groups, not just GL_n. The authors formulate a Serre weight conjecture asserting that, for p-restricted weights, this constraint is also sufficient, and they prove the elimination (necessity) direction.","feed_headline":"μ_p-symmetry constrains how crystalline Galois representations reduce mod p","feed_subtitle":"Prismatic cohomology reveals a hidden equivariance that forces inertial weights into a new partial order","key_machinery":"The argument has three load-bearing components. First, the prismatization W(k)^Δ and syntomification W(k)^Syn of W(k): a new map ρ†: Spf(S)/μ_p → W(k)^Δ is constructed that is congruent to the standard Breuil–Kisin prism map ρ_std modulo p but does not come from a prism structure integrally. Pulling back a prismatic F-crystal along ρ† produces the μ_p-equivariant modification N. Second, the geometry of μ_p-fixed points in affine Grassmannians: the orbits of L^+_p G on Gr^{μ_p} are indexed by dominant cocharacters, but their closure relations are governed by the ↑ partial order rather than the usual dominance order (Proposition 2.4.10, combining a combinatorial result on the Bruhat order with","core_discovery":"The central discovery is that crystallinity of a p-adic Galois representation imprints a μ_p-equivariant structure on the mod p reduction of its Breuil–Kisin module. This structure is invisible classically but emerges from prismatic cohomology: the Frobenius pullback of the Breuil–Kisin module is automatically μ_p-equivariant because the Frobenius endomorphism u ↦ u^p factors through the quotient by μ_p loop rotation. For crystalline representations, a canonical modification N of the Frobenius pullback — constructed from the Hodge filtration via the Rees module — is shown to be congruent to the Frobenius modification modulo p (Theorem 4.5.12, proved via the syntomification stack). This congr","pith_inferences":[],"forward_implications":["The constraint λ ↑ μ provides a uniform weight elimination result for Serre's conjecture for arbitrary unramified reductive groups, with no genericity assumptions — any Serre weight not satisfying the constraint cannot occur.","For GL_2 with Hodge–Tate weights (k,0) and p ≤ k ≤ 2p, the constraint recovers all four Berger–Breuil possibilities, confirming compatibility with known examples.","The μ_p-equivariant structure on mod p Breuil–Kisin modules is a new invariant that could distinguish reductions invisible to classical semisimplification, potentially sharpening local-global compatibility for automorphic forms.","The conjecture that W_cris(ρ) = W_explicit(ρ) = W(ρ) for p-restricted weights, if true, would give a complete and uniform Serre weight recipe for all unramified reductive groups.","The specialization argument via the p-Hecke stack (Corollary 2.7.4) provides a bridge between characteristic zero and characteristic p geometry that may apply to other mod p reduction problems."],"fun_headline_variants":["μ_p-equivariance constrains mod p reductions of crystalline Galois representations","Crystallinity imprints μ_p-structure on Breuil–Kisin reductions via prismatic cohomology","Hidden μ_p-symmetry in Breuil–Kisin modules constrains mod p inertial weights","New Serre weight conjecture from μ_p-equivariance of crystalline representations","Frobenius pullback of Breuil–Kisin modules carries natural μ_p-equivariance"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The load-bearing step is Theorem 4.5.12: that the canonical μ_p-equivariant modification of a Breuil–Kisin module, built from the Hodge filtration, agrees with the Frobenius modification modulo p. This is proved using the syntomification stack and a specific compatibility of pullbacks, and without it the μ_p-equivariant structure on the mod p reduction is not guaranteed.","fun_headline_variants_meta":{"raw":{"variants":["μ_p-equivariance constrains mod p reductions of crystalline Galois representations","Crystallinity imprints μ_p-structure on Breuil–Kisin reductions via prismatic cohomology","Hidden μ_p-symmetry in Breuil–Kisin modules constrains mod p inertial weights","New Serre weight conjecture from μ_p-equivariance of crystalline representations","Frobenius pullback of Breuil–Kisin modules carries natural μ_p-equivariance","Prismatic cohomology exposes μ_p-structure governing mod p reduction of crystalline reps","μ_p-fixed geometry of affine Grassmannians yields new inertial weight constraints"]},"model":"glm-5.2","effort":"low","cost_usd":0.0,"raw_usage":{"total_tokens":1144,"prompt_tokens":469,"completion_tokens":675,"prompt_tokens_details":null},"tokens_in":469,"tokens_out":675,"duration_ms":72816,"temperature":1.0,"reasoning_tokens":540,"cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-10T03:25:57.317159+00:00","model_set":{"reader":"glm-5.2"},"falsifier":"If there exists a crystalline representation whose Breuil–Kisin module's canonical modification (from the Hodge filtration) fails to be congruent to the Frobenius modification modulo p, then the μ_p-equivariant structure on M/pM is not established, the ↑ constraint on inertial weights need not hold, and the weight elimination result would fail for that representation.","supporting_citations":[],"review_version":1}