{"id":"47e29cce-b867-419a-b9b4-92860c6d69c7","arxiv_id":"1908.10733","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper establishes relative analogues of Hanke-Schick index obstruction theorems and shows that, under group assumptions, relative K-homology classes are almost flat modulo torsion.","lead":"This paper proves a relative version of the Hanke-Schick theorem, linking the Chang-Weinberger-Yu relative higher index to index pairings with almost flat vector bundles on manifolds with boundary. It also develops a quantitative version and identifies the range of the dual assembly map with almost flat classes for a class of residually amenable groups.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 3.5 rests on the unpublished relative almost monodromy correspondence (Theorem 2.24), quoted from [Kub19]; if its estimates fail, the relative Hanke–Schick nonvanishing conclusion collapses.","rationale":"The reader's weakest assumption is exactly the load-bearing point: Theorem 2.24 is used in both major applications (Theorem 3.5 and Theorem 5.13) and is quoted from an unpublished preprint. The rest of the proof of Theorem 3.5 is structurally coherent: the product construction, the quotient by the direct-sum ideal, and the exact-sequence argument for nonvanishing are all standard. The K-theory claim 'ker τ_* is identified with ⊕ K_0(A_n)' is imprecise as an isomorphism statement, but the needed conclusion follows from the six-term exact sequence and componentwise evaluation, so it is not a fatal flaw. A separate observation: Corollary 5.15 appears to have the inequality direction backwards, because ch = 0 would make the K-homology class torsion, so all index pairings vanish, contradicting infinite K-area as defined in Definition 2.21(2); this does not affect Theorem 3.5 but should be corrected. Since our main concern matches the reader's, the verdict remains CONDITIONAL.","tokens_in":39729,"tokens_out":24080,"duration_ms":233340,"concrete_test":"Obtain [Kub19] and verify the proof of [Kub19, Theorem 6.12] in full; specifically check that the maps α and β are continuous with the stated Lipschitz constants and that α sends 0-flat bundles to genuine (0-)representations. If the preprint is unavailable or contains a gap, attempt an independent proof using the absolute correspondence (2.15) plus the naturality of the relative mapping cone construction; failure to produce the claimed estimates would substantiate the concern.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The proof of Theorem 3.5 (Section 3.2) and the proof of Theorem 5.13 both invoke Theorem 2.24, the relative almost monodromy correspondence, quoted from the unpublished preprint [Kub19] (Definition 6.11, Theorem 6.12). In Theorem 3.5 the flat quotient bundle τ_*v is fed to the map α to obtain a stably relative representation α(τ_*v); the associated Kasparov bimodule Π is then paired with ℓ_{Γ,Λ} through Theorem 3.3 to identify ℓ_{Γ,Λ} ⊗ Π with [τ_*v]. The nonvanishing of α_{Γ,Λ}([M,N]) then follows. Every step in this chain depends on the existence of α and β with the asserted metric estimates and the constant C_am depending only on the open cover U. Since [Kub19] is not published and no proof is included here, the central claim rests on an unverified external result. If the estimates fail, or if α does not send 0-flat bundles to honest 0-representations, then Π is not a Kasparov bimodule of the type required by Theorem 3.3 and the equality ℓ_{Γ,Λ} ⊗ Π = [τ_*v] breaks. This would invalidate the relative Hanke–Schick obstruction (Theorem 3.5) and Corollary 3.6.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves a relative version of the Hanke–Schick nonvanishing theorem for the Chang–Weinberger–Yu relative higher index, together with quantitative index-pairing results and a dual-assembly statement. The main structural result is Theorem 3.3, which identifies the Kasparov product of the relative Mishchenko line element ℓ_{Γ,Λ} with a stably h-relative representation as the associated relative bundle. From this and the relative almost monodromy correspondence, Theorem 2.24, the author derives Theorem 3.5, asserting that infinite stably relative C*-K-area forces the relative higher index to be nonzero, and Corollary 3.6 gives the area-enlargeable boundary version. Section 4 reformulates Dadarlat's quasi-representation index theorem in quantitative K-theory, and Section 5 uses localization algebras to show that the range of the dual relative assembly map consists of stably almost flat K-theory classes, with Corollaries 5.14 and 5.15 as consequences.","tokens_in":40054,"tokens_out":13885,"duration_ms":153181,"significance":"If the results are correct, this is a meaningful extension of the Hanke–Schick obstruction from closed manifolds to manifolds with boundary, and it provides a quantitative relative index formula together with a new description of the range of the dual relative assembly map. The proof of Theorem 3.3 is detailed and appears to be a genuine structural contribution, and the reformulation of Connes–Gromov–Moscovici/Dadarlat in the framework of Oyono-Oyono–Yu quantitative K-theory is useful. However, the central nonvanishing theorems depend on the relative almost monodromy correspondence imported from the unpublished preprint [Kub19], and Corollary 5.15 as stated cannot be derived from the preceding modulo-torsion results, so the central claims are not yet fully verified by the manuscript alone.","major_comments":[{"comment":"The relative almost monodromy correspondence, Theorem 2.24, is quoted without proof from the unpublished preprint [Kub19], and it is load-bearing in two central places: in the proof of Theorem 3.5 the map α sends the flat quotient bundle τ_*v to a representation whose associated Kasparov bimodule Π is paired with ℓ_{Γ,Λ} through Theorem 3.3, and in Theorem 5.13 the same correspondence converts stable relative quasi-representations back to almost flat bundles. If any of the metric estimates in that correspondence fail, or if α does not send 0-flat bundles to exact 0-representations, then Π is not the required Kasparov bimodule and the equality ℓ_{Γ,Λ} ⊗ Π = [τ_*v] breaks. Since [Kub19] is not publicly available and no proof is supplied here, the central nonvanishing result rests on an unverified external result. The author should either include a proof of Theorem 2.24 or make the dependence explicit and conditional.","section":"§2.3, Theorem 2.24; §3.2, Theorem 3.5; §5.2, Theorem 5.13"},{"comment":"Corollary 5.15(1) states that (M,N) has infinite stably relative K-area if and only if ch(f_*[M,N]) = 0. This is not justified by the preceding results and, as stated, is incompatible with Definition 2.21(2). If ch(f_*[M,N]) = 0, then f_*[M,N] is torsion in K_0(BΓ,BΛ). For any integral K-theory class x, the index pairing ⟨x, f_*[M,N]⟩ is an integer, and if m f_*[M,N] = 0 then m⟨x, f_*[M,N]⟩ = 0, forcing the pairing to vanish. Hence a torsion fundamental class can never admit a nonzero index pairing with an integral almost flat class. The proof merely says the statement follows from Corollary 5.14, but Corollary 5.14 gives almost flatness only modulo torsion; one still has to produce a specific almost flat class with nonzero integer pairing. The intended statement is likely ch(f_*[M,N]) ≠ 0, and the passage from rational nonvanishing to a nonzero integer pairing requires an explicit argument clearing denominators.","section":"§5.2, Corollary 5.15"},{"comment":"In the proof of Theorem 3.5(1), the author concludes that τ_*(∏_n ⟨[v_n],[M,N]⟩) is nonzero by asserting that ker τ_* is identified with ⨁ K_0(A_n) through an isomorphism K_0(B) ≅ ∏ K_0(A_n). This is asserted without proof and is not a standard fact, since K-theory does not commute with arbitrary infinite products of C*-algebras; for example, K_0(∏ C) is C(βN,Z), not ∏ Z. Even if the special structure of the algebras B(P_n⊕Q_n) makes such an identification true, that would require a separate argument. Since the nonvanishing of the product class after quotienting by J is exactly what forces α_{Γ,Λ}([M,N]) to be nonzero, this is a load-bearing gap in the proof of Theorem 3.5(1).","section":"§3.2, proof of Theorem 3.5"}],"minor_comments":[{"comment":"There are numerous typographical artifacts from the source file, including 'Prelimilaries' in the Section 2 heading, 'mu∈I' in Section 2.3, and garbled diagrams with '/d47/' tokens in Lemma 3.9; these should be cleaned up before publication.","section":"Throughout"},{"comment":"The notation α^{δ,r}_{Γ,Λ}([M]) appears in a statement about a closed manifold M with no boundary pair; this should likely be α^{δ,r}_Γ([M]), and the subscript should be corrected for clarity.","section":"§4.2, Corollary 4.15"},{"comment":"The proof of Lemma 5.2 cites [DWW18, Proposition 4.3(a),(b)] for the vanishing of K-groups of three different quotients; it would be helpful to explicitly name which quotient is used for each asserted isomorphism, since the current display is terse.","section":"§5.1, Lemma 5.2"},{"comment":"In the definition of an (ε,U)-flat stably relative bundle, the phrase 'let T be a maximal subtree of the 1-skeleton N(U)' is used before the reader is told that T should restrict to a maximal subtree of N(U|_Y); please reorder the definition so that the compatibility condition is part of the stated hypothesis.","section":"§2.3, Definition 2.19"}],"recommendation":"major_revision","confidential_remarks":"The paper's central new statements are plausible and the author's approach is coherent, but the referee cannot verify Theorem 2.24 and Theorem 2.22 because they are quoted from the unpublished [Kub19]. In addition, Corollary 5.15 appears to contain a sign error or a missing argument, and the proof of Theorem 3.5 contains an unjustified K-theory product claim. These are fixable in principle, but the revision should either include the missing proofs or clearly restate the results as conditional on [Kub19]."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This paper proves the relative versions of Hanke–Schick and Dadarlat that the area has been expecting. The main theorems—Theorem 3.5 (relative higher index obstruction from infinite stably relative C*-K-area), Theorem 4.20 (relative quantitative index pairing), and Theorem 5.13 (range of the dual assembly map is stably almost flat)—are genuinely new relative statements, not just cosmetic generalizations. The proof strategy is sound: Theorem 3.3, which computes the Kasparov product of the relative Mishchenko line bundle with a stably h-relative representation, is the right tool, and the quantitative K-theory reformulation in Section 4 is careful and convincing.\n\nThe soft spots are real but not fatal. The biggest one is that the relative almost monodromy correspondence, Theorem 2.24, is quoted from the unpublished preprint [Kub19]. Both Theorem 3.5 and Theorem 5.13 rely on it for the maps α and β and their metric estimates. If those estimates fail, the constructions of the Kasparov bimodule Π and the flat bundle β(π) collapse. The author acknowledges the dependence, but the paper would be much stronger if the proof of Theorem 2.24 were included or the preprint made available. I don't see circularity—the companion paper is independent—but it is an external dependency.\n\nA smaller issue: Corollary 5.15 is stated as an iff via the Chern character, but it is derived from Corollary 5.14, which is a 'modulo torsion' statement. The leap from 'stably almost flat modulo torsion' to an actual equality against the Chern character needs more than the one-line \"immediately follows.\" This is likely fixable, but it needs spelling out.\n\nOverall, the mathematics is honest and the extensions are useful. The citation pattern is heavy on the author's own work, but that's appropriate when the prior papers are the right references. I would recommend sending this to a serious referee who can check the dependence on [Kub19] and the torsion-to-equality step in Section 5. The paper deserves engagement.","headline":"Relative Hanke–Schick and Dadarlat proved carefully, but the central theorem depends on an unpublished companion preprint.","tokens_in":40525,"tokens_out":2226,"would_cite":true,"duration_ms":23395,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["19K56","19K35","46L80","58J32"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that compact spin manifolds with boundary and infinite stably relative $C^*$-K-area have nonvanishing relative higher index, extending the closed-manifold almost-flat obstruction to manifolds with boundary.","keywords":["relative higher index","almost flat bundle","manifolds with boundary","positive scalar curvature","KK-theory","quantitative K-theory","dual assembly map"],"falsifier":"Construct, for $(\\Gamma,\\Lambda)=(\\mathbb{Z}^2,\\mathbb{Z})$ and the inclusion of the first factor, a sequence of stably relative quasi-representations $\\pi_n$ with flatness error $\\varepsilon_n=1/n$ whose associated bundles $\\beta(\\pi_n)$ are $\\varepsilon_n$-flat but for which $d(\\alpha\\circ\\beta(\\pi_n),\\pi_n)$ does not go to zero; this would violate the uniform error bound of Theorem 2.24, and with it the construction of $\\Pi$ in Theorem 3.5 collapses.","tokens_in":39555,"feed_emoji":"","tokens_out":12098,"duration_ms":114927,"temperature":0.7,"pith_summary":"This paper establishes a relative, boundary-version analogue of the classical almost-flat index obstruction: for a compact spin manifold $M$ with boundary $N$, if the pair has infinite stably relative $C^*$-K-area, then the relative higher index $\\mu_{\\Gamma,\\Lambda}([M,N])$ in $K_*(C^*(\\Gamma,\\Lambda))$ is nonzero. Here $\\Gamma=\\pi_1(M)$, $\\Lambda=\\pi_1(N)$, and \"infinite stably relative $C^*$-K-area\" means that at every flatness scale there is an almost flat stably relative bundle whose index pairing with $[M,N]$ is nonzero. Combined with a criterion from the companion preprint, this transfers the enlargeability obstruction from closed manifolds to manifolds with boundary and gives a direct index proof that certain manifolds with boundary admit no positive scalar curvature metric. The paper also proves a quantitative version in which the index pairing is computed through quantitative $K$-theory, and a dual theorem showing that, under mild group-pair assumptions, every element of the relative $K$-theory of classifying spaces is stably almost flat modulo torsion.","feed_headline":"Infinite relative K-area forces nonzero higher index","feed_subtitle":"Almost flat bundles on manifolds with boundary detect the relative index, extending the closed-manifold obstruction.","key_machinery":"The load-bearing object is the relative Mishchenko--Fomenko higher index map $\\alpha_{\\Gamma,\\Lambda}=\\mu_{\\Gamma,\\Lambda}^*$, defined as the Kasparov product with the relative Mishchenko line bundle $\\ell_{\\Gamma,\\Lambda}$, a KK-class built from the mapping cone C*-algebra of $C^*\\Lambda\\to C^*\\Gamma$ and the Mishchenko line bundles over $X$ and $Y$. The argument is carried by two mechanisms: Theorem 3.3, which converts Kasparov products with $\\ell_{\\Gamma,\\Lambda}$ into associated stably relative bundles, and the relative almost monodromy correspondence (Theorem 2.24), which converts almost flat stably relative bundles into stably relative quasi-representations and back with errors controlled by a universal constant. For the quantitative and dual results, the machinery is the algebraic relative higher index $\\alpha^{\\mathrm{alg}}_{\\Gamma,\\Lambda}$ valued in quantitative $K$-groups and the almost $*$-homomorphism induced by a stably relative quasi-representation.","core_discovery":"The central claim is Theorem 3.5: for a compact spin manifold $M$ with boundary $N$, writing $\\Gamma=\\pi_1(M)$, $\\Lambda=\\pi_1(N)$ with $\\Lambda\\to\\Gamma$ induced by inclusion, infinite stably relative $C^*$-K-area of $M$ implies $\\mu_{\\Gamma,\\Lambda}([M,N])\\neq 0$. The proof manufactures a single stably relative flat bundle over a quotient $D=B/J$ of a product algebra, whose fiberwise components are the assumed almost flat bundles, and applies the relative almost monodromy correspondence to obtain a stably h-relative representation $\\Pi$. Theorem 3.3, the paper's key structural step, identifies the Kasparov product $\\ell_{\\Gamma,\\Lambda}\\otimes_{C^*(\\Gamma,\\Lambda)}\\Pi$ with the associated stably relative bundle on $(M,N)$, so the pairing with $[M,N]$ is preserved and nonzero. A corollary is the relative enlargeability theorem, and a separate application upgrades the codimension-two index obstruction. Sections 4 and 5 respectively produce a quantitative index-pairing theorem and show that, under assumptions including the $\\gamma$-element and residual amenability, the dual relative higher index map is stably almost flat modulo torsion, giving a Chern-character criterion for infinite relative K-area.","pith_inferences":["Inference: if the relative almost monodromy correspondence of Theorem 2.24 is eventually published with explicit constants, the quantitative pairing theorem should yield effective scales: for a fixed finite CW pair, the smallest flatness error $\\varepsilon$ for which an $\\varepsilon$-flat stably relative bundle can detect a given K-homology class is controlled from below by a function of the relat","Inference: the dual theorem suggests a transfer principle: any future proof of rational surjectivity of the dual relative assembly map for a broader class of group pairs would automatically give stably almost flat representability for those pairs, because the KK-theoretic identification (Theorem 5.6) is independent of the specific group-pair hypotheses.","Inference: the codimension-two argument can be iterated: nonvanishing of the higher index on a codimension-two submanifold with trivial normal bundle and the stated $\\pi_1/\\pi_2$ conditions propagates to the ambient manifold, so the obstruction may detect positive scalar curvature in examples where the full higher index is hard to compute directly."],"forward_implications":["A compact spin manifold $M$ with collared boundary $N$ whose open end $M_\\infty$ is area-enlargeable has nonvanishing relative higher index $\\mu_{\\Gamma,\\Lambda}([M,N])$; this gives an enlargeability obstruction to positive scalar curvature on manifolds with boundary that needs no assembly-injectivity assumption.","When a codimension-two submanifold $N$ with $\\pi_1(N)\\to\\pi_1(M)$ injective, $\\pi_2(N)\\to\\pi_2(M)$ surjective, and trivial normal bundle has nonzero higher index, the relative higher index of the complement pair $(M_0,N_0)$ is also nonzero, so $M$ admits no positive scalar curvature metric.","If the relative higher index of a pair vanishes, then for sufficiently small flatness error every index pairing $\\langle[v],[M,N]\\rangle$ of an almost flat stably relative bundle $v$ vanishes, and the quantitative pairing theorem gives a local Chern-character formula for the pairing when a trace is present.","Under the group-pair assumptions (2.6), (2.7'), (2.8), and (5.8), every element of $K_0(B\\Gamma,B\\Lambda)$ is stably almost flat modulo torsion, so for a spin manifold with boundary the infinite stably relative K-area property is equivalent to vanishing of the relative Chern character $ch(f_*[M,N])$."],"supporting_citations":[{"why":"Defines the relative Mishchenko--Fomenko higher index as the Kasparov product with $\\ell_{\\Gamma,\\Lambda}$, proves duality of the dual map $\\beta_{\\Gamma,\\Lambda}$, and supplies the $\\gamma$-element surjectivity theorem used in Section 5.","marker":"[Kub18]"},{"why":"Provides the relative almost monodromy correspondence (Theorem 2.24) and the area-enlargeability-to-infinite-stably-relative-C*-K-area theorem (Theorem 2.22); both are quoted here and are load-bearing for Theorems 3.5 and 5.13.","marker":"[Kub19]"},{"why":"The closed-manifold enlargeability/nonvanishing theorem whose relative version is the paper's main goal; its product-bundle strategy is adapted in the proof of Theorem 3.5.","marker":"[HS06]"},{"why":"The infinite-covers companion to the closed-manifold obstruction, used as the model for the relative theorem alongside [HS06].","marker":"[HS07]"},{"why":"Supplies the non-relative almost monodromy maps and error constants (2.15) on which the relative correspondence of [Kub19] and the quantitative estimates build.","marker":"[CD18]"},{"why":"The Banach-KK quantitative index-pairing result that Section 4 reformulates in quantitative K-theory.","marker":"[Dad12]"},{"why":"Defines quantitative K-groups and the projection/unitary comparison lemmas used throughout Section 4.","marker":"[OOY15]"},{"why":"Shows almost flat representability of the image of the dual assembly map for residually amenable groups; the relative Theorem 5.13 follows its strategy.","marker":"[Dad14]"},{"why":"Quasi-diagonality of nuclear C*-algebras is used in Lemma 5.10 to produce finite-rank approximations of representations factoring through amenable quotients.","marker":"[TWW17]"}],"fun_headline_variants":["Relative index nonzero if K-area is infinite","Infinite relative K-area guarantees index obstruction","Almost flat bundles force nonzero relative index","Boundary relative index from infinite K-area"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is the quoted relative almost monodromy correspondence (Theorem 2.24): $\\varepsilon$-flat stably relative bundles and stably relative $\\varepsilon$-quasi-representations of $(\\Gamma,\\Lambda)$ are interchangeable with errors bounded by a fixed multiple of $\\varepsilon$; if those estimates fail, the product-bundle construction in Theorem 3.5 and the approximation in Theorem 5.13 have no foundation.","fun_headline_variants_meta":{"raw":{"variants":["Relative index nonzero if K-area is infinite","Infinite relative K-area guarantees index obstruction","Almost flat bundles force nonzero relative index","Boundary relative index from infinite K-area"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000189,"raw_usage":{"total_tokens":1315,"prompt_tokens":901,"completion_tokens":414,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":517,"completion_tokens_details":{"reasoning_tokens":360}},"tokens_in":517,"tokens_out":414,"duration_ms":4747,"temperature":1.0,"reasoning_tokens":360,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T10:35:51.755561+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct, for $(\\Gamma,\\Lambda)=(\\mathbb{Z}^2,\\mathbb{Z})$ and the inclusion of the first factor, a sequence of stably relative quasi-representations $\\pi_n$ with flatness error $\\varepsilon_n=1/n$ whose associated bundles $\\beta(\\pi_n)$ are $\\varepsilon_n$-flat but for which $d(\\alpha\\circ\\beta(\\pi_n),\\pi_n)$ does not go to zero; this would violate the uniform error bound of Theorem 2.24, and with it the construction of $\\Pi$ in Theorem 3.5 collapses.","supporting_citations":[],"review_version":1}