{"id":"6b3bad20-bc12-4214-9403-2fd7c48ec603","arxiv_id":"2607.27032","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Finite crepant covers of klt singularities scale normalized volume by the degree of the cover.","lead":"The paper proves a long-conjectured formula: when one singular space finitely covers another with matching log canonical classes, the size of the singularity (normalized volume) scales exactly by the number of sheets of the cover. This gives moduli and K-stability researchers a precise transfer rule for singularities under finite maps.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 4.1's δ-transfer relies on [LZ22, Thm 1.2(3)] having the exact quantitative statement; if it only gives K-semistability equivalence for Galois covers, the proof fails.","rationale":"The reader's weakest assumption precisely identifies the application of [LZ22, Theorem 1.2(3)] as the analytic input on which the proof of Theorem 4.1 depends. My review converges on the same point and makes it more specific: the proof does not merely need K-semistability to transfer, it needs a quantitative δ-invariant inequality to transfer through the non-Galois finite morphism h of Lemma 4.7. The manuscript gives no statement of [LZ22, Theorem 1.2], and the title of the cited paper indicates a focus on finite group actions, so there is a real risk that the theorem's hypotheses are not satisfied by h or that its conclusion is weaker than what the proof uses. This is not an internal inconsistency, but it is an unverified external assumption at a critical juncture. The reader's ACCEPT with moderate confidence is reasonable, but the central claim should be conditional on verification of [LZ22, Thm 1.2(3)]'s exact content. Hence I recommend CONDITIONAL rather than a full rejection.","tokens_in":17312,"tokens_out":29311,"duration_ms":378776,"concrete_test":"Inspect the statement of [LZ22, Theorem 1.2(3)] and verify: (a) its hypotheses allow arbitrary finite surjective morphisms h: E_X→E_Y with K_E_X+Δ_E_X = h^*(K_E_Y+Δ_E_Y), not just Galois quotients; (b) its conclusion includes the quantitative implication δ(E_X,Δ_E_X) < 1−ε ⇒ δ(E_Y,Δ_E_Y) < 1−ε (or equality of δ-invariants), not merely K-semistability equivalence. If either condition fails, the proof of Theorem 4.1 has a gap at that line.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"In the proof of Theorem 4.1, after finding a quasi-monomial v with A_X(v) < (1−2ϵ)S(ξ_X;v), the author obtains δ(E(ξ),Δ_E(ξ)) < 1−ϵ for quasi-regular ξ near ξ_X (Lemma 4.4). The next step is the load-bearing one: 'Combining this with Lemma 4.7 and [LZ22, Theorem 1.2(3)], we obtain δ(E(ξ'),Δ_E(ξ'))<1−ϵ for any quasi-regular ξ'∈q_*(U_ϵ).' This requires [LZ22, Theorem 1.2(3)] to assert a quantitative δ-invariant inequality for the finite log-crepant morphism h: E(ξ_X)→E(ξ_Y) constructed in Lemma 4.7 — not merely an if-and-only-if of K-semistability, and not only for Galois covers. The manuscript never states the content of [LZ22, Theorem 1.2]; the title of [LZ22] ('Equivariant K-stability under finite group action') suggests it may cover only quotients by finite groups. Lemma 4.7's h is not generally Galois. If [LZ22, Thm 1.2(3)] does not give the exact δ-transfer, then Theorem 4.1 — and therefore the central equality in Theorem 1.1 — is not established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves the finite degree formula for normalized volumes: for a finite surjective morphism f:(x∈(X,Δ_X))→(y∈(Y,Δ_Y)) between klt singularities with K_X+Δ_X=f^*(K_Y+Δ_Y), one has cvol(x,X,Δ_X)=deg(f)cvol(y,Y,Δ_Y). This confirms a conjecture of Liu–Li–Xu, Zhuang, and Xu–Zhuang. The proof combines stable degeneration with the theory of log Fano cones: after lifting a special degeneration from Y to X and lifting the torus action, it transfers K-semistability via a local analogue of [LZ22, Theorem 1.2], and then computes volumes by a graded Hilbert-function count (Lemma 5.1). The reverse inequality follows from a divisorial valuation argument.","tokens_in":17687,"tokens_out":13245,"duration_ms":184426,"significance":"If valid, this is a significant result: it removes the Galois assumption from the finite degree formula, a basic tool in the local stability theory of singularities. The proof is synthetic, with no free parameters; the main volume computation is a clean homological count. The main caveat is that the proof depends on the precise quantitative statement of [LZ22, Theorem 1.2], which is not stated in the paper; the referee cannot verify the crucial δ-transfer step from the manuscript alone.","major_comments":[{"comment":"The step 'Combining this with Lemma 4.7 and [LZ22, Theorem 1.2(3)]' is load-bearing and is not justified as written. The manuscript never states the content of [LZ22, Theorem 1.2] nor its hypotheses. In particular, the finite morphism h:E(ξ_X)→E(ξ_Y) from Lemma 4.7 is not Galois in general, while the title of [LZ22] suggests a Galois/group-quotient statement. If [LZ22, Theorem 1.2(3)] does not give the quantitative δ-transfer for non-Galois finite log-crepant morphisms, the K-semistability transfer in Theorem 4.1, and hence the proof of Theorem 1.1, fails. Please quote the theorem and verify h satisfies all hypotheses.","section":"§4, proof of Theorem 4.1"},{"comment":"The final sentence 'The remaining implication is completely the same' is not immediate: the morphism in Theorem 4.1 goes from X to Y, not from Y to X. The forward direction uses [LZ22] to turn δ(E(ξ_X))<1−ϵ into δ(E(ξ_Y))<1−ϵ; the reverse direction would need the opposite transfer or an additional argument. Since Theorem 1.1 only needs the forward direction, the statement of Theorem 4.1 should either be restricted to that direction, or the reverse implication should be proved.","section":"§4, Theorem 4.1, reverse implication"}],"minor_comments":[{"comment":"The reduction 'Using [Sta26, 02LN], we may assume f^{-1}({y})={x} as sets' is too quick. For a finite surjective morphism between normal varieties, the sum of local degrees over the fiber of a closed point equals deg(f) only under additional hypotheses (e.g., flatness). Please supply the argument or prove the corollary by summing the local statement over the finitely many preimages.","section":"§1, Corollary 1.2"},{"comment":"The phrase 'quasi-regular Reeb vectors are dense' should be accompanied by a reference, and the intersection argument with U_ϵ and the neighborhood from Lemma 4.5 should be spelled out.","section":"§4, proof of Theorem 4.1"},{"comment":"The choice of homogeneous elements b_i with weights χ_i forming a C(Y)-basis of C(X) should be justified, and the claim that dim Supp(Q)≤n−1 should be stated explicitly.","section":"§5, Lemma 5.1"},{"comment":"The introduction says 'we prove a local analog of [LZ22, Theorem 1.2(1)]', while the proof of Theorem 4.1 uses [LZ22, Theorem 1.2(3)]; please clarify which part is used.","section":"§1"}],"recommendation":"major_revision","confidential_remarks":"The referee's recommendation is driven by the need to verify [LZ22, Theorem 1.2]. The author should be asked to quote the theorem verbatim and state precisely which part is used. The reverse implication of Theorem 4.1 is also in need of clarification, but the central finite degree formula only requires the forward direction."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Read it. This looks like the real proof of the finite degree conjecture, not a repackaging of the Galois case. The stable-degeneration strategy is genuinely different from [XZ21]: lift the special degeneration, lift the log Fano cone structure, transfer local K-semistability, then count Hilbert functions. The volume equality in Lemma 5.1 is the cleanest part—once you have the lifting and the K-semistability transfer, the normalized volumes match by a graded count with a torsion term that dies for dimension reasons. The lifting lemmas in Section 3 are also careful and largely self-contained; Lemma 3.2's etale-fundamental-group argument is a credible way to lift torus actions.\n\nThe soft spot is exactly the one you flagged. Theorem 4.1 is load-bearing, and its proof uses [LZ22, Theorem 1.2(3)] without stating what that theorem says. The morphism h in Lemma 4.7 is finite and log-crepant but not generally Galois, and [LZ22] is titled \"Equivariant K-stability under finite group action.\" If Theorem 1.2(3) is only the quotient/Galois statement, the δ-transfer from E(ξ) to E(ξ') does not follow, and Theorem 4.1 loses its grip. The author explicitly says the proof relies on an analytic input and that no algebraic proof is known; that is honest, but it also means the central result is not self-contained. I do not see an internal contradiction, and the dependency is not circularity—it is a dependency on exactly the kind of quantitative statement that must be verified statement-by-statement. If [LZ22] does cover this non-Galois setting, the concern evaporates; if not, the proof needs a new argument at that step.\n\nMinor issues: Proposition 3.1 does a finite base change and then asserts deg(g0)=deg(f); that is plausible after the ramification control, but a referee should check it. The reduction in Corollary 1.2 to a single preimage is standard. Nothing else wobbles.\n\nWho this is for: anyone working on normalized volumes, local K-stability, or boundedness/ACC applications. It deserves a serious referee. The referee's main job is to open [LZ22] and confirm whether Theorem 1.2(3) really gives a δ-inequality transfer for non-Galois finite log-crepant maps between log Fano pairs. If yes, accept with minor comments; if no, the author needs to fill that gap before the result is established.","headline":"A serious, genuinely new proof of the non-Galois finite degree formula; the one thing to check before believing it is whether [LZ22, Thm 1.2(3)] really gives the non-Galois δ-transfer the proof needs.","tokens_in":18117,"tokens_out":3537,"would_cite":true,"duration_ms":53484,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14B05","14J17","13A18","14J45"],"pacs":[],"model":"deepseek-v4-flash","headline":"For any finite log-crepant cover of klt singularities, normalized volumes scale exactly by the degree.","keywords":["normalized volume","klt singularity","K-semistability","log Fano cone","finite morphism","stable degeneration","degree formula","log pair"],"falsifier":"Compute both sides of the identity for an explicit non-Galois finite log-crepant cover, such as a degree-3 cover of an A_2 singularity branched along a torus-invariant divisor, and check that the normalized volume ratio equals the degree; a mismatch would refute the theorem. Alternatively, exhibit isogeny-equivariant log Fano cones related by a finite log-crepant morphism where one is K-semistable and the other is not, which would contradict Theorem 4.1.","tokens_in":17230,"feed_emoji":"📐","tokens_out":4242,"duration_ms":62273,"temperature":0.7,"pith_summary":"The paper proves the finite degree formula for normalized volumes: if a finite surjective morphism between klt singularities pulls back the log canonical divisor, then the normalized volume of the source singularity equals the degree of the map times the normalized volume of the target. The formula was previously known for Galois covers and had been conjectured in full generality. The proof degenerates both singularities to log Fano cones, lifts the degeneration and the torus action along the cover, transfers K-semistability through an isogeny-equivariant finite morphism, and computes the volume contribution exactly. This settles a central conjecture in the local stability theory of singularities and gives a powerful tool for computing normalized volumes.","feed_headline":"Finite covers scale a singularity's normalized volume exactly by degree","feed_subtitle":"A conjecture connecting local volumes of singularities under finite maps is proved by degenerating to K-semistable cones.","key_machinery":"The normalized volume cvol is the infimum of (log discrepancy)^dim times volume over valuations centered at the singular point; its minimizer induces a stable degeneration to a K-semistable log Fano cone. The proof's load-bearing steps are the lifting of the special degeneration along the finite cover, the lifting of the torus action and log Fano cone structure, the K-semistability transfer under isogeny-equivariant finite log-crepant morphisms, and the exact degree-multiplicativity of the weighted volume on log Fano cones.","core_discovery":"Theorem 1.1 states that for a finite surjective morphism f between klt singularities with K_X+Δ_X = f^*(K_Y+Δ_Y), the normalized volumes satisfy cvol(x,X,Δ_X)=deg(f)·cvol(y,Y,Δ_Y). The proof reduces the general case to log Fano cones via stable degeneration, lifts the degeneration to the cover, shows the lifted central fiber is K-semistable whenever the target is (Theorem 4.1, via approximation by quasi-regular Reeb vectors and an analytic transfer result), and then computes the normalized volumes of the vertices via a graded length comparison (Lemma 5.1). Lower semicontinuity gives one inequality and a direct volume estimate gives the reverse.","pith_inferences":["The proof's reliance on a cited analytic input suggests that finding an algebraic replacement for the quasi-regular approximation step would make the whole argument purely algebraic, which the paper explicitly leaves open.","The volume identity in Lemma 5.1 may be interpretable as a Riemann–Roch statement for filtered algebras, hinting at a general framework where normalized volumes behave like degrees of finite extensions of graded rings.","One could expect the degree formula to hold for all quasi-monomial valuations, not only minimizers, which would strengthen the numerical control over finite log-crepant morphisms.","The equivariant lifting construction (Lemma 3.3) might extend to profinite or reducible coverings, giving a route to degree formulas for non-finite but quasi-finite log-crepant maps."],"forward_implications":["The formula gives a local analogue of Riemann–Hurwitz: normalized volume behaves multiplicatively under finite log-crepant covers, enabling explicit computations for covers of known singularities.","Applications that previously required Galois covers, such as boundedness and moduli statements, now work for arbitrary finite morphisms between klt singularities.","Theorem 4.1 provides a general method to produce new K-semistable log Fano cones from old ones by taking finite covers with compatible Reeb vectors.","The lifted degeneration construction shows that stable degenerations are compatible with finite morphisms up to base change, refining the interaction between singularities and their degenerations.","The two-sided bound obtained in the proof gives a practical numerical criterion for comparing normalized volumes under finite maps, independent of knowing minimizers explicitly."],"fun_headline_variants":["Finite covers multiply singularity volumes by degree","Exact degree scaling for normalized volumes under finite maps","Klt singularities: finite maps scale volumes by degree","Normalized volume formula: finite covers and degree","Finite morphisms: volume scales exactly by degree"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The whole argument leans on a cited analytic theorem about equivariant K-semistability under finite group actions, applied to Kollár components obtained by quasi-regular approximation; if that theorem does not apply at the approximation step, the K-semistability transfer fails and the main equality collapses.","fun_headline_variants_meta":{"raw":{"variants":["Finite covers multiply singularity volumes by degree","Exact degree scaling for normalized volumes under finite maps","Klt singularities: finite maps scale volumes by degree","Normalized volume formula: finite covers and degree","Finite morphisms: volume scales exactly by degree"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000138,"raw_usage":{"total_tokens":933,"prompt_tokens":631,"completion_tokens":302,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":375,"completion_tokens_details":{"reasoning_tokens":228}},"tokens_in":375,"tokens_out":302,"duration_ms":9304,"temperature":1.0,"reasoning_tokens":228,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T03:16:45.706926+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute both sides of the identity for an explicit non-Galois finite log-crepant cover, such as a degree-3 cover of an A_2 singularity branched along a torus-invariant divisor, and check that the normalized volume ratio equals the degree; a mismatch would refute the theorem. Alternatively, exhibit isogeny-equivariant log Fano cones related by a finite log-crepant morphism where one is K-semistable and the other is not, which would contradict Theorem 4.1.","supporting_citations":[],"review_version":2}