{"id":"f273dad9-8ae4-43fc-8e1d-6c17248b1244","arxiv_id":"2506.24059","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For spin manifolds with iterated conical singularities, scalar-mean curvature comparison forces equality and rigidity, and nonnegative scalar curvature implies nonnegative ADM mass.","lead":"This paper proves a rigidity theorem for spaces with cone-like singularities: if their scalar and mean curvature are at least as large as a standard model, then equality must hold. It also proves a positive mass theorem for such singular spaces.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 3.6's spectral-gap proof implicitly normalizes φ(1)=1; without that normalization the lower bound |P| ≥ √((n-1)(n-2))/2 can fail, undermining the Fredholmness argument that the whole comparison theorem depends on.","rationale":"The reader's verdict is CONDITIONAL, with weakest_assumption identified as the codimension-at-least-three condition. I agree that the paper is plausible and the strategy coherent, but I find a more specific and more load-bearing gap in Lemma 3.6. This lemma is the hinge of the Fredholmness proof: it supplies the spectral gap needed for the Brüning–Seeley self-adjointness criterion. The algebraic step in its proof appears to require φ(1)=1 without stating a normalization; moreover, the normalization is not obviously compatible with the global hypotheses, especially the boundary metric coincidence in Case (1). A concrete parameter regime (φ(1)=1/2, flat torus domain link) shows the claimed lower bound can fail, so the issue is not merely cosmetic. This does not move the overall verdict because the gap may be repairable by a corrected normalization or a sharper argument, but it sharpens the reason the paper should remain CONDITIONAL rather than ACCEPT. The reader's rationale mentioned Lemma 3.6 only briefly; my concern is more central than the codimension condition, which the paper itself motivates with a counterexample.","tokens_in":35692,"tokens_out":44512,"duration_ms":441586,"concrete_test":"Independently re-derive Lemma 3.6 without assuming φ(1)=1. Then compute the lowest eigenvalue of P² for the explicit cone map with n=3, domain link a flat torus, target link the round 2-sphere with Sc=4, φ(1)=1/2, φ′(1)=10, and f a map with ||df||=1 at the relevant point. If the smallest eigenvalue of P² is less than (n−1)(n−2)/4 = 1/2, the lemma is false as stated; if it is ≥1/2 for all such choices, then the gap is only in the written proof and can be repaired.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central proof relies on Lemma 3.6 to obtain the Witt-type spectral gap |P| ≥ √((n-1)(n-2))/2 for the link operator P, which is what makes the twisted Dirac operator essentially self-adjoint and Fredholm on the cone (Section 3.2). But the algebraic chain in Lemma 3.6 is not valid as written. At the link r=1, the comparison condition Sc_ChatL ≥ ||dF||² F*Sc_CL reads Sc_hatL − A ≥ max{φ′(1)², φ(1)²||df||²}(f*Sc_L − A/φ(1)²), where A=(n−1)(n−2). The proof instead uses an inequality of the form Sc_hatL − A ≥ max{φ′(1), φ(1)||df||}²(f*Sc_L − A), which is justified only if φ(1)=1. No such normalization is stated, and it is not invariant under target metric rescaling when the boundary metric coincidence condition (Case (1)) is imposed. The subsequent curvature-term estimate in P² then fails in explicit configurations: take n=3, a flat torus domain link (Sc_hatL=0, A=2), a round target link with Sc_L=4 (so f*Sc_L−A/φ(1)² = −4 if φ(1)=1/2), φ(1)=1/2 and φ′(1)=10. The comparison inequality holds at the link, but the curvature term in P² is only 0.25, below A/4=0.5. Thus the claimed spectral gap (and hence self-adjointness/Fredholmness) is unsupported. Since every later step—dichotomy, rigidity, positive mass—uses this Fredholmness, this is the most load-bearing gap in the paper.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves a scalar-mean curvature comparison theorem for compact spin manifolds with iterated asymptotically conical singularities (IACS). The main result, Theorem 5.6, asserts that if the domain and target satisfy the scalar and mean curvature comparison inequalities, the target has nonnegative curvature operator and nonnegative boundary second fundamental form, and all singular links have dimension at least two, then the two inequalities are equalities; additional rigidity conclusions hold when the target has positive Ricci curvature or is flat. The proof uses a twisted Dirac operator, a Witt-type spectral-gap estimate for the link operator to obtain Fredholmness, and a dichotomy argument that avoids explicit computation of the index. Applications include a rigidity theorem for Euclidean domains and a spin positive mass theorem for asymptotically flat IACS manifolds.","tokens_in":36081,"tokens_out":26301,"duration_ms":293846,"significance":"If the proof is completed, this is a substantial extension of Llarull/Lott-type comparison rigidity to spaces with iterated conical singularities. The paper has real strengths: the codimension-two counterexample in Remark 2.4 correctly identifies why the condition on links is needed; the proof has no fitted parameters or post-hoc adjustments; and the dichotomy argument is an interesting way to bypass eta-invariant computations for the index. The Hardy-type estimates and Rellich compactness for IACS spaces in Section 5 are useful technical contributions. However, several load-bearing steps are only sketched, so the current version is not yet a complete proof of the advertised theorems.","major_comments":[{"comment":"The inequality chain in Lemma 3.6 reads Sc_hatL - A >= max{phi'(1)^2, phi(1)^2||df||^2} (f^*Sc_L - A)/phi(1)^2 >= ||df||^2 (f^*Sc_L - A). The second inequality is not valid for arbitrary sign of f^*Sc_L - A; it requires f^*Sc_L - A to be nonnegative. This does follow from the hypothesis that the curvature operator of the target cone is nonnegative: by Proposition 2.6, R_CL|_{wedge^2 TL} = (1/r^2)(R_L - Id), so R_L >= Id and hence Sc_L >= (n-1)(n-2). The manuscript omits this sign argument, and without it the algebra is wrong because multiplying by a larger positive factor reverses an inequality when the bracket is negative. I also note that a normalization phi(1)=1 is not needed: the factor 1/phi(1)^2 is exactly the pullback scalar curvature, and the inequality max{...}^2 >= phi(1)^2||df||^2 supplies the correct constant.","section":"Section 3.2, Lemma 3.6"},{"comment":"The Kato-Rellich perturbation step that passes from the model operator D_{\\tilde V^R} to the original twisted Dirac operator is only sketched. The paper asserts that the error terms have asymptotic order o(1/gamma) or c(hat b)/gamma with c(hat b) -> 0, and that 'taking R to be very big and shrinking hat U and U to be very small' gives the required relative bound with b < 1. These assertions are load-bearing for self-adjointness, Fredholmness, and therefore all subsequent conclusions. The manuscript should provide the actual estimates, the explicit dependence on R and the neighborhood sizes, and a verification that the cited results from [5] apply uniformly to the family of metrics considered; as written this is a proof sketch rather than a proof.","section":"Section 3.3, after Eq. (3.4)"},{"comment":"In Case (2) the proof asserts that the 'classical Atiyah-Singer index theorem applied to the boundary' yields a nonzero section in the kernel of D^{\\partial,\\partial} when chi(partial M) != 0 and F has nonzero degree. The operator D^{\\partial,\\partial} is a twisted Dirac operator on S(T hat M \\oplus F^*TM)|_{\\partial hat M}, and the dependence of its index on the boundary map F_\\partial is not written down. The paper should state the relevant index formula, identify the coefficient bundle and grading, and explain how the degree of F_\\partial enters; otherwise the dichotomy in Case (2) is unsupported.","section":"Section 4, Theorem 4.1, Case (2)"}],"minor_comments":[{"comment":"The title and abstract contain typographical artifacts ('CUR V ATURE', 'cur-vature') that should be corrected.","section":"Title/Abstract"},{"comment":"After defining A = (n-1)(n-2), the proof should explicitly state that f^*Sc_L - A >= 0 as a consequence of nonnegative curvature operator of the target cone, so that the inequality max{...}^2 >= ||df||^2 can be multiplied by a nonnegative quantity.","section":"Section 3.2, Lemma 3.6"},{"comment":"In the Cauchy-Schwarz step of the Hardy-type estimate, the bound on \\int |\\sigma|^2 r^{l-1} dr in terms of \\|1/r \\sigma\\|^2 uses that r is bounded away from infinity in the conical neighborhood; this localization should be stated explicitly.","section":"Section 3.1, Proposition 3.2"},{"comment":"The sentence 'It is not hard to see that ||d f_{\\hat b_0}|| is bounded above by some constant' should be justified, for example by compactness of the link and smoothness of the family f(hat b, .); this boundedness is needed for the induction.","section":"Section 5.3"},{"comment":"The sentence 'By definition, the singularities of M lie entirely within the interior. Therefore, the map F_\\partial extends to a smooth map F : M -> Omega' is terse; although the extension is immediate because Omega is a ball, the collar construction should be described for completeness.","section":"Section 6.1, Theorem 6.1"}],"recommendation":"major_revision","confidential_remarks":"The stress-test concern about a missing phi(1)=1 normalization in Lemma 3.6 does not land on the manuscript as stated: the factor 1/phi(1)^2 is part of the pullback scalar curvature, and the algebraic step only needs f^*Sc_L - A >= 0, which follows from nonnegative target curvature operator. The real problem in that lemma is that this sign condition is never stated. The more serious issues are the sketched perturbation estimates in Section 3.3 and the unproved index assertion in Case (2) of Theorem 4.1. These are fixable but require real work, hence major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear Colleague,\n\nYou should know about this paper because it takes the scalar-mean comparison theorems of Lott and Cecchini-Zeidler and the spin positive mass theorem to manifolds with iterated conical singularities. The main theorems are plausible, and the strategy is the right kind of ambitious: define FACS and IACS cleanly, set up a twisted Dirac operator, prove Fredholmness via a spectral gap on the link, then use a dichotomy argument to avoid computing an index with eta invariants. The dichotomy trick is a genuine innovation, and Remark 2.4 is honest about the codimension-two obstruction.\n\nThe problem is Lemma 3.6, the spectral-gap lemma. It is the load-bearing step for self-adjointness and Fredholmness, and its proof has a gap. At the link, the comparison condition gives something like Sc_hatL - A >= max{phi'(1)^2, phi(1)^2 ||df||^2} (f^*Sc_L - A/phi(1)^2). The proof then drops the phi(1) factors and asserts this is at least ||df||^2 (f^*Sc_L - A). That is valid only with an unstated normalization phi(1)=1, and there is no reason such a normalization exists or is preserved by target metric rescaling. The stress-test example with a flat torus link and a constant link map is enough to show the local inequality at the link does not rule out a zero eigenvalue for P, so the claimed lower bound |P| >= sqrt((n-1)(n-2))/2 is unsupported. The global comparison hypothesis may well rule out such maps, but the proof of Lemma 3.6 never uses it. As written, the analytic core of the paper is not established.\n\nThe rest of the paper has smaller soft spots: the Kato-Rellich perturbation in Section 3.3 is asserted rather than proved, and the induction for iterated singularities in Section 5 leans on the same spectral gap. The rigidity arguments in Sections 4 and 6 follow the established pattern and are fine if the Fredholmness is granted.\n\nMy take: this is a real paper with a real idea, not a desk-reject. It should go to referees, because the main theorems are important and the gap may be repairable. A referee should insist on a corrected Lemma 3.6 before acceptance.\n\nBest","headline":"A serious and important extension of scalar-mean comparison to iterated conical singularities, but the spectral-gap lemma that carries the analytic core has a proof gap that needs to be fixed.","tokens_in":36601,"tokens_out":11164,"would_cite":false,"duration_ms":106219,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["53C21","53C27","58J20","53C24"],"pacs":[],"model":"deepseek-v4-flash","headline":"Scalar-mean rigidity extends to spin manifolds with iterated conical singularities.","keywords":["scalar-mean curvature comparison","iterated conical singularities","twisted Dirac operator","Fredholmness","rigidity","positive mass theorem","spin manifolds"],"falsifier":"Take the cylinder example described in Remark 2.4: \\(\\hat M=[0,1]\\times $S^{1}$/\\sim\\) with \\(\\hat g=$dr^{2}$+$r^{2}$g_{$S^{1}$}\\), \\(M=[0,a]\\times $S^{1}$/\\sim\\) with \\(g=$dr^{2}$+$a^{{-2}}$$r^{2}$g_{$S^{1}$}\\), and \\(F(r,x)=(ar,x)\\); for \\(a>1\\) this satisfies the comparison hypotheses but not the equality conclusions. Producing such a one-dimensional-link pair in a compact boundary-preserving or nonzero-degree setting would directly disprove Theorem 5.6 as stated.","tokens_in":35484,"feed_emoji":"📐","tokens_out":6879,"duration_ms":71855,"temperature":0.7,"pith_summary":"The paper establishes a scalar-mean curvature comparison theorem for compact spin manifolds with iterated asymptotically conical singularities (IACS), spaces whose singular strata are recursively built from cones over lower-dimensional singular links. The claim is that if an asymptotically conical map between two such manifolds satisfies the comparison inequalities \\(\\mathrm{Sc}_{\\hat g}\\ge \\|dF\\|^2F^*\\mathrm{Sc}_g\\) and \\(H_{\\hat g}\\ge \\|d(F_\\partial)\\|\\)\\((F_\\partial)^*H_g\\), with the target having nonnegative curvature operator and nonnegative boundary second fundamental form, then both inequalities are equalities. Equality then drives rigidity: a target with positive Ricci curvature makes the map a Riemannian covering up to homothety, and a flat target forces a Ricci-flat domain. The proof runs a twisted Dirac operator with a spectral boundary condition and uses a dichotomy argument so the Fredholm index never has to be computed explicitly. The same machinery yields a rigidity theorem for Euclidean domains and a spin positive mass theorem for asymptotically flat manifolds with iterated conical singularities, so scalar-mean rigidity survives non-isolated, iterated singularities.","feed_headline":"Spin rigidity theorem extends to iterated conical singularities","feed_subtitle":"A Dirac dichotomy yields scalar-mean comparison equalities, a Euclidean-domain rigidity theorem, and a positive mass theorem.","key_machinery":"The carrying object is the twisted Dirac operator \\(D\\) on the spinor bundle of \\(T\\hat M\\oplus F^*TM\\), studied with the absolute boundary condition \\(\\sqrt{-1}\\hat c(\\hat\\nu)c(\\nu)\\$\\sigma$=-\\$\\sigma$\\) on \\(\\partial\\hat M\\). The argument first proves self-adjointness and Fredholmness: on each conical neighborhood the operator reduces to a model that is self-adjoint when the link operator \\(P\\) has spectral gap \\(|P|\\ge \\tfrac12\\sqrt{(n-1)(n-2)}\\), and this gap follows from the scalar-curvature comparison and nonnegative curvature operator. A Hardy-type inequality and a compact Sobolev embedding, both requiring links of dimension at least two, convert self-adjointness into Fredholmness. The proof is completed by a dichotomy: if the boundary-value Dirac operator has a kernel section, the kernel section directly forces the comparison equalities; if it is invertible, an index-theoretic input on the boundary produces a harmonic section, which again forces the equalities. This avoids computing the index of the singular twisted Dirac operator.","core_discovery":"The central result, Theorem 5.6, the precise form of Theorem 1.1, asserts that for compact spin \\(C^m\\)-manifolds with iterated conical singularities and smooth boundaries, links of dimension at least two, nonnegative target curvature operator, and nonnegative boundary second fundamental form, the hypotheses \\(\\mathrm{Sc}_{\\hat g}\\ge \\|dF\\|^2F^*\\mathrm{Sc}_g\\) and \\(H_{\\hat g}\\ge \\|d(F_\\partial)\\|\\)\\((F_\\partial)^*H_g\\) force equality in both inequalities, provided either the boundaries coincide as Riemannian manifolds or the boundary has nonzero Euler characteristic and \\(F\\) has nonzero degree. If \\(\\mathrm{Ric}_g>0\\), equality forces \\(\\|dF\\|\\) to be constant and \\(F:(\\hat M,a\\hat g)\\to(M,g)\\) to be a Riemannian covering map; if \\(M\\) is flat, then \\(\\hat M\\) is Ricci flat. The theorem is proved first for fiberwise asymptotically conical singularities and then extended to iterated conical singularities by induction on stratum depth, with Fredholmness of the twisted Dirac operator as the load-bearing analytic step.","pith_inferences":["The dichotomy argument suggests that the Fredholm index of the singular twisted Dirac operator is not needed for rigidity, so structurally similar comparison statements could hold for other elliptic operators once Fredholmness is established.","The sharpness discussion indicates that codimension-two singularities genuinely break the result; one testable direction is whether a weaker Hardy-type inequality or a different boundary condition can restore a limiting version of the theorem for codimension-two strata.","One could try to extend the comparison statement beyond spin by adapting the paper's estimates to minimal-surface or \\(\\mu\\)-bubble methods, which have produced smooth low-dimensional comparison theorems without spin.","The positive mass theorem part suggests constructing explicit asymptotically flat IACS examples with two-dimensional links to probe whether the zero-mass flatness conclusion is sharp."],"forward_implications":["If Theorem 5.6 is correct, no asymptotically conical map satisfying the comparison inequalities can improve scalar curvature or mean curvature strictly; the inequalities are always saturated.","In the positive-Ricci target case, the map is a Riemannian covering up to a constant homothety, so the domain metric is determined by the target metric up to scale.","A flat target forces the domain to be Ricci flat, giving a singular-space analogue of flat rigidity.","Theorem 6.1 asserts that a compact spin IACS manifold with nonnegative scalar curvature and a boundary map of nonzero degree to a closed convex Euclidean hypersurface is flat under the stated boundary conditions.","Theorem 6.4 asserts nonnegative ADM mass for asymptotically flat spin IACS manifolds with nonnegative scalar curvature and links of dimension at least two, with zero mass forcing flatness."],"supporting_citations":[{"why":"Supplies the smooth scalar-mean comparison theorem whose statement and Dirac-operator strategy are generalized to singular manifolds.","marker":"[24]"},{"why":"Provides the sharp scalar curvature comparison theorem for closed manifolds that the boundary case builds on.","marker":"[23]"},{"why":"Gives the scalar curvature estimate for compact symmetric spaces that the comparison framework extends.","marker":"[16]"},{"why":"Shows how boundary Dirac terms can be controlled in a scalar-mean comparison, informing the boundary condition setup.","marker":"[10]"},{"why":"Supplies the self-adjointness criterion for regular singular first-order operators used to prove Fredholmness on cones.","marker":"[7]"},{"why":"Establishes Dirac operator analysis on conical singularities that underlies the link spectral-gap argument.","marker":"[12]"},{"why":"Provides the foundational treatment of spectral geometry of singular Riemannian spaces that motivates the singular-manifold setup.","marker":"[11]"},{"why":"Defines the ADM mass and provides the boundary spinor computation used in the positive mass theorem.","marker":"[3]"},{"why":"Provides the harmonic-spinor dichotomy method adapted here to avoid computing the index.","marker":"[33]"}],"fun_headline_variants":["Inequality forces rigidity in iterated conical singularities","Dirac dichotomy proves scalar-mean comparison for singular spaces","Positive mass theorem holds for iterated conical singularities","Rigidity and positive mass for spin manifolds with cone singularities","Comparison theorem for iterated conical singularities"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument depends on the singular strata having codimension at least three, equivalently all links having dimension at least two; if a one-dimensional link appears, the Hardy inequality, compact embedding, and spectral-gap bound fail, and the paper itself notes a codimension-two counterexample.","fun_headline_variants_meta":{"raw":{"variants":["Inequality forces rigidity in iterated conical singularities","Dirac dichotomy proves scalar-mean comparison for singular spaces","Positive mass theorem holds for iterated conical singularities","Rigidity and positive mass for spin manifolds with cone singularities","Comparison theorem for iterated conical singularities"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000254,"raw_usage":{"total_tokens":1527,"prompt_tokens":864,"completion_tokens":663,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":480,"completion_tokens_details":{"reasoning_tokens":583}},"tokens_in":480,"tokens_out":663,"duration_ms":6601,"temperature":1.0,"reasoning_tokens":583,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T21:26:05.219689+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the cylinder example described in Remark 2.4: \\(\\hat M=[0,1]\\times $S^{1}$/\\sim\\) with \\(\\hat g=$dr^{2}$+$r^{2}$g_{$S^{1}$}\\), \\(M=[0,a]\\times $S^{1}$/\\sim\\) with \\(g=$dr^{2}$+$a^{{-2}}$$r^{2}$g_{$S^{1}$}\\), and \\(F(r,x)=(ar,x)\\); for \\(a>1\\) this satisfies the comparison hypotheses but not the equality conclusions. Producing such a one-dimensional-link pair in a compact boundary-preserving or nonzero-degree setting would directly disprove Theorem 5.6 as stated.","supporting_citations":[{"cited_title":"Index theory for scalar curvature on manifolds with boundary","cited_arxiv_id":null,"evidence_quote":"Supplies the smooth scalar-mean comparison theorem whose statement and Dirac-operator strategy are generalized to singular manifolds."},{"cited_title":"Sharp estimates and the Dirac operator","cited_arxiv_id":null,"evidence_quote":"Provides the sharp scalar curvature comparison theorem for closed manifolds that the boundary case builds on."},{"cited_title":"Goette and U","cited_arxiv_id":null,"evidence_quote":"Gives the scalar curvature estimate for compact symmetric spaces that the comparison framework extends."},{"cited_title":"Scalar and mean curvature comparison via the Dirac operator","cited_arxiv_id":null,"evidence_quote":"Shows how boundary Dirac terms can be controlled in a scalar-mean comparison, informing the boundary condition setup."},{"cited_title":"An index theorem for first order regular singular op- erators","cited_arxiv_id":null,"evidence_quote":"Supplies the self-adjointness criterion for regular singular first-order operators used to prove Fredholmness on cones."},{"cited_title":"The Dirac operator on spaces with conical singularities and posi- tive scalar curvatures","cited_arxiv_id":null,"evidence_quote":"Establishes Dirac operator analysis on conical singularities that underlies the link spectral-gap argument."},{"cited_title":"Spectral geometry of singular Riemannian spaces","cited_arxiv_id":null,"evidence_quote":"Provides the foundational treatment of spectral geometry of singular Riemannian spaces that motivates the singular-manifold setup."},{"cited_title":"The mass of an asymptotically flat manifold","cited_arxiv_id":null,"evidence_quote":"Defines the ADM mass and provides the boundary spinor computation used in the positive mass theorem."},{"cited_title":"A new proof of the positive energy theorem","cited_arxiv_id":null,"evidence_quote":"Provides the harmonic-spinor dichotomy method adapted here to avoid computing the index."}],"review_version":1}