{"id":"4e274d6b-7b08-4435-ad90-6db1e1f1b51a","arxiv_id":"1908.01329","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"A new groupoid model for Elek's C*-algebras is constructed, and the C*-algebra is shown to be nuclear exactly when the associated Schreier graph has local property A.","lead":"This paper recasts Elek's C*-algebras associated with uniformly recurrent subgroups as reduced groupoid C*-algebras of an explicitly constructed etale groupoid. The reformulation yields simpler proofs and adds the missing converse in the nuclearity characterization.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The density proof in Theorem 3.3 uses the false bound ||Σ_{|η|≤R} h||_2^2 ≤ (|Q|+1)^R ||h||_2^2; a Z-graph example gives left side ≈9 with R=1 while the right side is 3.","rationale":"I read the paper in good faith and checked the groupoid construction that the reader flagged. I did not find a fatal flaw there: the source map is well-defined through minimal representatives, the topology is the projective-limit topology, and the étale and continuity arguments are plausible. The most concrete and load-bearing problem I found is in the proof of Theorem 3.3, the central isomorphism theorem. The density argument that CZ is dense in Cc(G) in the reduced norm relies on the estimate ||Σ_{l(η)≤R} h||_2^2 ≤ (|Q|+1)^R ||h||_2^2. This estimate is false: the translation sum operator has norm up to the ball size N, so the squared norm can be N^2 ||h||^2. My Γ=Z counterexample is a clean, minimal computation that settles the point. Because the density argument still works with the corrected constant, the main theorem is not overturned; the paper needs a small but necessary correction. Therefore I recommend CONDITIONAL acceptance rather than rejecting or leaving the verdict unchanged. I disagree with the reader's identification of the weakest assumption: the groupoid construction, while intricate, appears sound, whereas the norm-bound error is explicit and verifiable.","tokens_in":12553,"tokens_out":32104,"duration_ms":319650,"concrete_test":"Recompute the key density estimate for Γ=Z, Q={±1}, H={0}, S=Z, with R=1. Take h_K ∈ l2(Z) defined by h_K(n)=1/√(K+1) for 0≤n≤K and 0 elsewhere. Take f∈Cc(G) to be the locally constant function equal to 1 on arrows ([[x]],η) with d(x,ηx)≤1 and 0 otherwise. Compute ||π_{[[H]]}(f)h_K||_2^2 = Σ_x (Σ_{|y-x|≤1} h_K(y))^2 = (9K-1)/(K+1). For K large this is ≈9, while the paper's bound gives (|Q|+1)^R ||f||_∞^2 ||h_K||_2^2 = 3. This demonstrates the inequality in the proof of Theorem 3.3 is wrong; the corrected Cauchy-Schwarz estimate gives the constant (|Q|+1)^{2R}=9.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"In the proof of Theorem 3.3 the author needs to show that CZ is dense in Cc(G) in the reduced norm. The argument passes through the displayed estimate ||π_{[[H]]}(f)h||_2^2 ≤ ||f||_∞^2 ||Σ_{η:l(η)≤R} h||_2^2 ≤ (|Q|+1)^R ||f||_∞^2 ||h||_2^2. The final inequality is false as stated. For fixed h, the operator T_Rh = Σ_{l(η)≤R} h_η has l2-operator norm at most the ball size N=(|Q|+1)^R, so the squared norm can be as large as N^2 ||h||^2, not N ||h||^2. A concrete failure occurs for Γ=Z, Q={±1}, H={0}, S=Z, R=1. Let h be the normalized indicator of {0,1,...,K} and let f be the indicator of all arrows with d(x,ηx)≤1. Then ||π(f)h||_2^2 = (9K-1)/(K+1) → 9, while the claimed bound is 3||h||^2=3. Thus the proof of Theorem 3.3 contains an incorrect numerical estimate. The conclusion is not destroyed, because replacing (|Q|+1)^R by (|Q|+1)^{2R} makes the approximation argument still work: ||f-f_ε||_r ≤ C_R ||f-f_ε||_∞ → 0. So the central isomorphism is very likely true, but the proof as written needs repair.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper recasts Elek's construction of C*-algebras associated with uniformly recurrent subgroups (URSs) in the language of groupoid C*-algebras. For a URS Z of a finitely generated discrete group Γ, the author constructs an ample, minimal, Hausdorff étale groupoid G whose unit space is homeomorphic to Z, built from inverse limits of finite sets of rooted, labelled balls in a Schreier graph. The main theorem (Theorem 3.3) identifies Elek's reduced C*-algebra C*_r(Z) with the reduced groupoid C*-algebra C*_r(G), with the isomorphism extending the natural bijection between local kernels and locally constant compactly supported functions on G. Section 4 shows that G is a continuous, open quotient of the transformation groupoid Z⋊Γ. Section 5 gives groupoid proofs of Elek's simplicity result for generic URSs and proves a full characterization: C*_r(Z) is nuclear if and only if the Schreier graph of any H∈Z has local property A (Theorem 5.3 and Corollary 5.4).","tokens_in":12906,"tokens_out":11119,"duration_ms":116587,"significance":"The groupoid picture is a natural and potentially useful reformulation of Elek's algebras. It makes Elek's algebras accessible to standard tools of groupoid C*-algebras, explains the simplicity and nuclearity results conceptually, and adds the converse implication in the nuclearity characterization, thereby strengthening Elek's Theorem 8. The main construction is elegant and the paper is mostly clearly written. The proof of Theorem 3.3 contains an incorrect norm estimate, but the error is local and fixable; the central claims are very likely correct. The paper offers no machine-checked proofs or executable code, but the mathematical arguments are explicit enough to be checked by hand.","major_comments":[{"comment":"In the proof of Theorem 3.3, near the end, the displayed estimate reads ||π_{[[H]]}(f)h||_2^2 ≤ ||f||_∞^2 ||Σ_{η: l(η)≤R} h||_2^2 ≤ (|Q|+1)^R ||f||_∞^2 ||h||_2^2. The final inequality is false. The operator T_R = Σ_{η: l(η)≤R} λ_η is a sum of at most (|Q|+1)^R unitaries, so ||T_R|| ≤ (|Q|+1)^R, and therefore ||T_R h||_2^2 is bounded by ((|Q|+1)^R)^2 ||h||_2^2, not by (|Q|+1)^R ||h||_2^2. A concrete failure occurs for Γ=Z with Q={±1}, R=1, and h the normalized indicator of {0,1,...,K}: ||T_R h||_2^2 = (9K-1)/(K+1) → 9 while the claimed bound is 3. The subsequent approximation argument still works if the exponent is changed to 2R (equivalently, if the factor is replaced by (|Q|+1)^{2R}), since only convergence of the approximation is needed. Thus the conclusion of Theorem 3.3 is not in question, but the proof as written is invalid at this step and must be repaired.","section":"Theorem 3.3, density estimate"}],"minor_comments":[{"comment":"The sentence 'All elements of Gx are therefore represented by We employ this description of the space Z...' is a grammatical fragment and should be rewritten.","section":"Section 3, first paragraph"},{"comment":"The phrase 'for γ whose lengths as words in the generators exceeds n' has a subject-verb agreement error and should read 'exceed n'.","section":"Section 3, construction of F_n"},{"comment":"In the computation of ||ρ^n_{γH}||_2^2, the summation is written over γ'H∈G/H; this should be Γ/H.","section":"Theorem 5.3, converse direction"},{"comment":"After the norm estimates, the inequality ||ρ^n_{γH}-ρ^n_{γ'H}||_2^2 ≤ 1/N yields only ||ρ^n_{γH}-ρ^n_{γ'H}||_2 ≤ N^{-1/2}; a small relabelling (for example, requesting 1/N^2) is needed to match the definition of local property A. This is harmless but should be stated explicitly.","section":"Theorem 5.3, local property A conclusion"},{"comment":"The abstract and some body text contain typographical errors (e.g., 'it s', 'o f', 'a nd', 'characterisation o f nuclearity'); a careful proofreading pass is advised.","section":"General"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe short version: this is a genuinely useful groupoid reformulation of Elek's construction, and the new converse on nuclearity is a real result. But the proof of Theorem 3.3 has a wrong constant in a norm estimate; the gap is repairable but as written the density argument doesn't go through.\n\nWhat is new: the groupoid G built as an inverse limit of ball-isomorphism classes, the description of G as a quotient of the transformation groupoid Z ⋊ Γ, and the amenability equivalence in Theorem 5.3. The construction is genuinely from first principles; there is no circularity. The paper uses standard groupoid theorems as tools, and the overall structure is clear. The converse of Elek's nuclearity result (Corollary 5.4) is a true extension.\n\nThe soft spot: in the density argument for Theorem 3.3, the author asserts ||Σ_{l(η)≤R} h_η||_2^2 ≤ (|Q|+1)^R ||h||_2^2. That is false. The operator T_Rh = Σ_{l(η)≤R} h_η has l2 norm at most the number of group elements of length ≤ R, so the squared norm can be as large as (|Q|+1)^{2R} ||h||^2. A concrete Z example (Q={±1}, R=1, h the normalized indicator of {0,...,K}) gives left side tending to 9 while the right side is 3. So as written, the bound does not prove what it needs to. The good news: the argument is repairable by replacing R with 2R in the exponent, and the conclusion of the theorem is very likely correct. Still, the proof needs fixing.\n\nThere are also several compressed justifications—'easy to see', 'easy to check'—and one garbled sentence in the construction of G. These are minor, but they add friction.\n\nThe central isomorphism and the nuclearity characterization appear robust despite the estimate problem. This is not a case where the whole idea fails; it's a case where a proof has a fixable flaw.\n\nWho benefits: anyone working on C*-algebras from URSs, on Schreier graph property A, or on groupoid C*-algebra techniques. I'd bring it to a reading group.\n\nRecommendation: send it to peer review. A referee should require the norm estimate correction and a few expansions, but the paper deserves referee time.","headline":"A useful groupoid reformulation with a real new converse, but the key density proof has a repairable norm-estimate error.","tokens_in":13415,"tokens_out":4555,"would_cite":true,"duration_ms":41409,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["46L05","22A22","20F65","37B05"],"pacs":[],"model":"deepseek-v4-flash","headline":"Elek's uniformly-recurrent-subgroup algebras are canonically isomorphic to reduced groupoid C*-algebras, and nuclearity is equivalent to local property A of the Schreier graph.","keywords":["uniformly recurrent subgroups","Elek algebras","groupoid C*-algebras","étale groupoids","Schreier graphs","local property A","nuclearity","simplicity"],"falsifier":"Find a uniformly recurrent subgroup for which the projective-limit source map $s(x,\\gamma)=\\gamma x$ is discontinuous, or a URS whose Schreier graph lacks local property A but whose associated reduced C*-algebra is nuclear; either would refute Theorem 3.3 or Corollary 5.4.","tokens_in":12350,"feed_emoji":"","tokens_out":14310,"duration_ms":128763,"temperature":0.7,"pith_summary":"This paper shows that the C*-algebras Elek associated with uniformly recurrent subgroups are, in a canonical way, reduced groupoid C*-algebras. A uniformly recurrent subgroup (URS) is a minimal closed conjugation-invariant subspace of the space of subgroups of a finitely generated group; to each one Elek had attached an algebra built from local kernels on a Schreier graph. The paper constructs an ample minimal Hausdorff étale groupoid $\\mathcal{G}$ whose unit space is homeomorphic to the URS $Z$, and proves that its reduced C*-algebra $C^*_r(\\mathcal{G})$ is canonically isomorphic to Elek's $C^*_r(Z)$ (Theorem 3.3). The gain is that analytic questions become graph-geometric ones: the Schreier graph has local property A exactly when $\\mathcal{G}$ is topologically amenable, which is exactly when $C^*_r(Z)$ is nuclear (Corollary 5.4). A reader should care because Elek's algebras are a flexible source of simple C*-algebras with unusual traces, and the groupoid picture makes their nuclearity and simplicity readable from the combinatorics of a single Schreier graph.","feed_headline":"Elek algebras are groupoid algebras; nuclearity is graph-detectable","feed_subtitle":"Every Elek algebra is a reduced groupoid C*-algebra; local property A of the Schreier graph settles nuclearity.","key_machinery":"The central object is the projective-limit groupoid $\\mathcal{G} = \\lim_{\\leftarrow} F_n \\setminus \\{\\infty\\}$, where each level $F_n$ records the $n$-ball around a representative of each root-label isomorphism class in $E_n$, plus a point $\\infty_n$ for 'too far away'. An arrow $(x, \\gamma)$ has range $x$, source $\\gamma x$ defined coordinatewise from a minimal-length representative, and composition is $(y,\\gamma')(x,\\gamma) = (y,\\gamma\\gamma')$ when $x = \\gamma' y$. The topology is the subspace topology of the projective limit, and the unit space $\\lim_{\\leftarrow} E_n$ is homeomorphic to the URS $Z$. This machinery carries the argument by converting a local kernel $K \\in C_Z$ of width $N$ into the locally constant compactly supported function $f_K(x,\\gamma) = K(x_M,\\gamma x_M)$ on $\\mathcal{G}$; the conversion is a bijection with locally constant functions, preserves convolution and involution, and extends to a canonical isomorphism of reduced C*-algebras.","core_discovery":"For a finitely generated group $\\Gamma$ and a uniformly recurrent subgroup $Z \\subset \\mathrm{Sub}(\\Gamma)$, fix $H \\in Z$ and its Schreier graph $S = S^Q_\\Gamma(H)$ with respect to a finite symmetric generating set $Q$. The paper constructs a groupoid $\\mathcal{G}$ as a projective limit of the finite sets $F_n = \\bigsqcup_{[x_n] \\in E_n} B_n(S, x_n) \\sqcup \\{\\infty_n\\}$, where $E_n$ is the set of root-label isomorphism classes of $n$-balls in $S$; removing the limit point $\\infty$ gives the arrows. Theorem 3.3 states that the reduced C*-algebra $C^*_r(Z)$ built from local kernels on $S$ is canonically isomorphic to the reduced groupoid C*-algebra $C^*_r(\\mathcal{G})$, and that $\\mathcal{G}^{(0)} \\cong Z$. The groupoid is ample, minimal, Hausdorff and étale, and is a quotient of the transformation groupoid $Z \\rtimes \\Gamma$ by identifying group elements that move the root to the same vertex inside a ball (Proposition 4.1). From this, Corollary 5.4 gives the full equivalence: $S^Q_\\Gamma(H)$ has local property A if and only if $\\mathcal{G}$ is topologically amenable if and only if $C^*_r(Z)$ is nuclear. In addition, when $Z$ is generic, $\\mathcal{G}$ is principal and $C^*_r(Z)$ is simple (Proposition 5.1, Corollary 5.2).","pith_inferences":["The inverse-limit construction does not need a group action: it would attach an Elek-type C*-algebra to any uniformly recurrent rooted graph, giving a purely combinatorial route to new groupoid C*-algebras.","Because $\\mathcal{G}$ is a quotient of $Z \\rtimes \\Gamma$, one could test whether $C^*_r(\\mathcal{G})$ is Morita equivalent to a crossed product by a partial action of $\\Gamma$ on $Z$; if so, K-theory of Elek algebras would be computable from URS dynamics.","The equivalence with local property A suggests an algorithmic criterion: nuclearity can be decided by checking, at every scale, whether vertices with isomorphic balls admit approximately orthogonal localizing functions, a finite-ball condition in a uniformly recurrent graph."],"forward_implications":["Every Elek algebra $C^*_r(Z)$ is the reduced C*-algebra of an ample minimal Hausdorff étale groupoid, so the groupoid toolkit applies to it: ideals, traces, K-theory, and amenability can be read from $\\mathcal{G}$.","Nuclearity of $C^*_r(Z)$ is equivalent to local property A of the Schreier graph of any subgroup in the URS, upgrading Elek's earlier one-direction result to an if-and-only-if criterion.","If $Z$ is generic, $\\mathcal{G}$ is principal and $C^*_r(Z)$ is simple; in the amenable case the converse also holds, so under local property A, simplicity of $C^*_r(Z)$ is equivalent to genericity of $Z$.","The groupoid $\\mathcal{G}$ is a quotient of the transformation groupoid $Z \\rtimes \\Gamma$, with range fibre $\\Gamma/H$ instead of $\\Gamma$, so the Elek algebra records how group elements move the root within finite balls rather than the full conjugation action."],"supporting_citations":[{"why":"Defines the local kernel algebra CZ and its reduced C*-algebra C*_r(Z), the object the paper recasts as a groupoid algebra.","marker":"[2]"},{"why":"Identifies the inverse limit of ball-isomorphism classes with the URS Z, which becomes the unit space of the groupoid.","marker":"[2, Lemma 6.1.4]"},{"why":"Supplies the ball-isomorphism characterization of generic URSs used to prove G is principal when Z is generic.","marker":"[2, Proposition 2.3]"},{"why":"Provides the theorem that a minimal principal étale groupoid has simple reduced C*-algebra, used for Corollary 5.2.","marker":"[6, Proposition 4.3.7]"},{"why":"Supplies the equivalent characterization of topological amenability used in the converse direction of Theorem 5.3.","marker":"[1, Proposition 2.2.13]"},{"why":"States that an étale groupoid has nuclear reduced C*-algebra exactly when it is topologically amenable, used for Corollary 5.4.","marker":"[1, Corollary 6.2.14]"},{"why":"Gives the original definition of topological amenability against which the local-property-A functions are checked in Theorem 5.3.","marker":"[5, page 92]"}],"fun_headline_variants":["Elek algebras are groupoid C*-algebras; nuclearity iff local property A","Groupoid realization of Elek algebras: nuclearity from Schreier graph property A","Every Elek algebra is a reduced groupoid algebra; nuclearity decided by Schreier graph","Nuclearity of Elek algebras characterized by Schreier graph local property A"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the inverse limit built from Schreier-graph ball classes, with the source map sending a path to its endpoint, is a genuine topological groupoid with continuous structure maps; if the endpoint map is not continuous, the isomorphism $C^*_r(Z) \\cong C^*_r(\\mathcal{G})$ and every downstream conclusion collapses.","fun_headline_variants_meta":{"raw":{"variants":["Elek algebras are groupoid C*-algebras; nuclearity iff local property A","Groupoid realization of Elek algebras: nuclearity from Schreier graph property A","Every Elek algebra is a reduced groupoid algebra; nuclearity decided by Schreier graph","Nuclearity of Elek algebras characterized by Schreier graph local property A"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000405,"raw_usage":{"total_tokens":2102,"prompt_tokens":934,"completion_tokens":1168,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":550,"completion_tokens_details":{"reasoning_tokens":1080}},"tokens_in":550,"tokens_out":1168,"duration_ms":9451,"temperature":1.0,"reasoning_tokens":1080,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:15:45.020815+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find a uniformly recurrent subgroup for which the projective-limit source map $s(x,\\gamma)=\\gamma x$ is discontinuous, or a URS whose Schreier graph lacks local property A but whose associated reduced C*-algebra is nuclear; either would refute Theorem 3.3 or Corollary 5.4.","supporting_citations":[{"cited_title":"Elek, Uniformly recurrent subgroups and simple C∗-algebras, J","cited_arxiv_id":null,"evidence_quote":"Defines the local kernel algebra CZ and its reduced C*-algebra C*_r(Z), the object the paper recasts as a groupoid algebra."}],"review_version":1}