{"id":"cf298542-53b5-4d85-851b-676747b3a5a1","arxiv_id":"2501.04709","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"In a unital separable C*-algebra, a set G is hyperrigid exactly when every sequence of representations that converges weakly on G converges strongly on the whole algebra.","lead":"The paper shows that for separable C*-algebras, hyperrigidity can be checked using only representations, not all completely positive maps. This gives a simpler, purely representation-theoretic test for whether a set of operators controls the whole algebra under approximation.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 2.1's proof falsely assumes every representation of a separable C*-algebra has separable Hilbert space; nonseparable H is not covered.","rationale":"The central claim—that hyperrigidity for separable A is equivalent to WOT-to-SOT convergence for representations—is plausible, and the proof strategy (cut a UCP map via Stinespring, build approximating representations with a shift, apply the representation-only condition) is standard. I checked the shift construction and equation (2.3); the algebra appears correct for separable H. The most load-bearing flaw is the assertion that dim H ≤ ℵ0. This is not a minor typo: it is used three times to pass to separable H, K, and K⊖H. Since condition (ii) in Theorem 2.1 explicitly quantifies over every Hilbert space, the proof as written does not establish the stated equivalence. I do not see a counterexample; the gap is likely closed by the standard invariant-subspace reduction. Because the reader's verdict was already CONDITIONAL, and this concern reinforces rather than overturns it, I keep the verdict unchanged. The reader's weakest assumption (separability of A) is related but distinct: our concern is about the Hilbert space, not the C*-algebra.","tokens_in":5492,"tokens_out":14295,"duration_ms":125142,"concrete_test":"Implement the proposed reduction: let π, π_n on H fail condition (ii) with witnessing vector h and element a. Let H0 be the closed linear span of all vectors π_{i_1}(b_1)...π_{i_k}(b_k)h with i_j ∈ N∪{0} (π_0 = π) and b_j ∈ A. Show H0 is separable (A is separable), invariant under every π_n and π, and that weak-operator convergence on G persists when restricted to H0 while strong-operator convergence on A fails at h. If this works, the false 'dim H ≤ ℵ0' statement is a harmless defect and Theorem 2.1 can be repaired; if the reduction cannot be carried out, the theorem lacks a proof for nonseparable H.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In the proof of (ii)⇒(i) in Theorem 2.1, the authors assert: 'We always have that dim H ⩽ ℵ0 as a consequence of the separability of A.' This is false. A separable C*-algebra admits nonseparable representations; e.g., let A = C([0,1]) and define π(f) = diag(f(t)) on ℓ²([0,1]). Hence the reductions to dim H = ℵ0, dim K = ℵ0, and dim(K⊖H) = ℵ0 do not cover arbitrary H, even though condition (ii) quantifies over every Hilbert space. The subsequent shift construction and the appeal to condition (ii) on K therefore apply only to separable H. A standard repair would be to show that a failure of (ii) on a nonseparable H yields a failure on a separable common invariant subspace (the closed span of words in π_n and π applied to a witnessing vector). The paper does not supply this argument. Thus the central equivalence is not fully proven as written; it is a gap, not a counterexample, and it is likely repairable.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies hyperrigidity for subsets of unital C*-algebras and proposes, in Theorem 2.1, a characterization of hyperrigidity solely in terms of representations of the algebra, without reference to completely positive maps. Specifically, for a nonempty subset G of a unital separable C*-algebra A, the paper claims that G is hyperrigid if and only if, whenever representations π_n and π on the same Hilbert space satisfy weak-operator convergence of π_n(g) to π(g) for every g in G, they also satisfy strong-operator convergence of π_n(a) to π(a) for every a in A. The proof of the forward direction uses the known characterizations in Theorem 1.2, while the reverse direction uses Stinespring dilation and a unilateral shift construction to convert a putative UCP extension into a sequence of representations.","tokens_in":5664,"tokens_out":9572,"duration_ms":92787,"significance":"If Theorem 2.1 is correct, it is a clean and useful reformulation of hyperrigidity: it eliminates UCP maps entirely from the definition and reduces the checking to a representation-theoretic approximation condition. The proof strategy is standard and builds on established results (Arveson's unique extension property characterization, Kleski's WOT characterization, and Stinespring dilation), and there is no indication of circularity. The paper is a contribution to the ongoing study of hyperrigidity and its characterizations. However, as written, the proof of the reverse implication contains a gap concerning nonseparable Hilbert spaces, and the presentation has several typos that obscure the central argument. The core idea appears sound and the gap is likely repairable, but the manuscript needs substantive revision before it can be accepted.","major_comments":[{"comment":"The assertion 'We always have that dim H ⩽ ℵ0 as a consequence of the separability of A' is false: a separable C*-algebra admits nonseparable representations, for example A = C([0,1]) acting by multiplication on ℓ²([0,1]). Because of this, the reductions 'without loss of generality dim H = ℵ0', 'dim K = ℵ0', and 'dim(K ⊖ H) = ℵ0' do not cover arbitrary Hilbert spaces, even though condition (ii) quantifies over every Hilbert space. The shift construction and the subsequent appeal to condition (ii) on K are therefore justified only in the separable case. The manuscript should supply a reduction from a nonseparable H to a separable common invariant subspace (for instance, for a fixed vector x, the closed span of {π_n(a)x, π(a)x : n ∈ N, a ∈ A}), or otherwise handle the nonseparable case explicitly.","section":"§2, proof of Theorem 2.1, (ii)⇒(i)"},{"comment":"The proof invokes 'Theorem 1.2(v)', but Theorem 1.2 has no condition (v); the intended reference appears to be condition (ii) of Theorem 2.1 itself. In the same passage, Eq. (2.3) writes 'w-lim_{n→∞} Φ_n(a)' although the objects just defined are the representations π_m and the limit should be as m → ∞. These errors are not merely typographical: they concern the exact step in which condition (ii) is applied, so they must be corrected for the argument to be readable and verifiable.","section":"§2, after the definition of π_m, Eq. (2.3)–(2.4)"}],"minor_comments":[{"comment":"The sentence 'Suppose dim H < ℵ0 ... we infer from the ℵ0-version of the proof' refers to a case that has not yet been established at that point in the proof. This is not circular if the finite-dimensional case is treated after the ℵ0 case is proved, but the order of exposition should be made explicit.","section":"§2, proof of Theorem 2.1, finite-dimensional reduction"},{"comment":"The notation '¯A0 = A with cardA0 ≤ ℵ0, and ¯H0 = H with cardH0 ≤ ℵ0' is confusing; it should say that A0 and H0 are countable dense subsets of A and H respectively. This matters because the separability of K_min depends on choosing such sets.","section":"§2, definition of K_min"},{"comment":"The definition of σ∞ is ambiguous: it should be a representation on K ⊖ H, but the text says 'Define the representation σ∞ : A → B(K) by σ∞ = σ ⊕ σ ⊕ ...' with K decomposed as H ⊕ (K ⊖ H). Clarify how σ∞ acts on the first summand or define it only on K ⊖ H and extend by zero on H.","section":"§2, definition of σ∞"},{"comment":"There are several spelling and grammar errors: 'faithfull' should be 'faithful', 'without loos of generality' should be 'without loss of generality', 'In view of beging of the proof' should be 'In view of the beginning of the proof', and Theorem 2.1 states 'every representations' instead of 'every representation'.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The central theorem is plausible and the proof strategy is recognizable from the dilation literature, but the false claim about separability of representation spaces is a real gap in the written proof. The paper should not be accepted until that gap is closed or the theorem is restricted to separable Hilbert spaces. The typos in the proof, especially the nonexistent 'Theorem 1.2(v)' and the Φ_n/π_m mix-up in Eq. (2.3), should also be fixed before resubmission."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's the one thing you should know: Theorem 2.1 is a genuinely useful reformulation of hyperrigidity, giving a characterization that quantifies only over representations, not UCP maps. But the proof as written has a real gap—it asserts that every representation of a separable C*-algebra lives on a separable Hilbert space, which is false. The universal representation of any infinite-dimensional separable C*-algebra is nonseparable, for instance.\n\nWhat's new: the theorem says that for a unital separable A and nonempty G ⊂ A, G is hyperrigid iff for every Hilbert space H and every sequence of representations π_n, weak convergence on G implies strong convergence on A. That's genuinely different from Arveson's and Kleski's characterizations, which quantify over UCP maps or use the unique extension property. If it holds, it gives a cleaner way to check hyperrigidity in the separable case. The proof idea—turning a UCP map into a weak limit of representations via Stinespring and a shift—is clever and shows the authors know the machinery.\n\nThe soft spots: the step I flagged is not a typo; it's a false mathematical claim. The proof reduces to dim H = ℵ0 using that claim, so condition (ii) is never applied to nonseparable H, even though it is stated for every H. That leaves the central equivalence unproven as written. A standard cyclic-subspace repair should work: given a witnessing vector where strong convergence fails, take the closed span of all words in π_n and π applied to that vector. This subspace is separable and invariant under all the maps, so you get a failure on a separable H. But the paper doesn't supply that argument. I see this as a repairable gap, not a fatal flaw—the fix is short and probably standard. Separately, there are typos: a reference to \"Theorem 1.2(v)\" which doesn't exist, and Φ_n/π_n confusion in the limit calculations. These are minor and don't affect the ideas.\n\nBottom line: the result is a nice reformulation for the hyperrigidity community, and the authors clearly engage seriously with the existing literature. But the proof is incomplete as written, so I wouldn't cite it as a theorem yet. It deserves a serious referee, and with a revision that properly handles nonseparable H—and cleans up the typos—it should be acceptable.","headline":"Useful representation-only characterization of hyperrigidity, but the proof has a real gap: separable C*-algebras can have nonseparable representations, and the main theorem is incomplete as written, though likely repairable.","tokens_in":6204,"tokens_out":5265,"would_cite":false,"duration_ms":50596,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["46G10","47B15","47A63","44A60"],"pacs":[],"model":"deepseek-v4-flash","headline":"In separable C*-algebras, hyperrigidity is the same as a weak-to-strong convergence property for representations.","keywords":["hyperrigidity","completely positive maps","representations","weak operator topology","strong operator topology","unital C*-algebras","Stinespring dilation","approximation theory"],"falsifier":"Find a unital C*-algebra A and a nonempty subset G such that every sequence of representations that converges weakly on G converges strongly on A, yet some sequence of unital completely positive maps converges weakly on G without converging strongly on all of A; in the separable case the theorem predicts such a pair cannot exist, so any concrete example would refute the characterization.","tokens_in":5280,"feed_emoji":"","tokens_out":7329,"duration_ms":67648,"temperature":0.7,"pith_summary":"Every unital C*-algebra has a notion of hyperrigidity: a set G is hyperrigid when the asymptotic behavior of any sequence of unital completely positive maps is fixed by what it does on G. This paper proves that, for separable algebras, this property can be checked using only representations, which are maps that preserve multiplication, rather than all completely positive maps. Concretely, G is hyperrigid exactly when any sequence of representations that converges weakly on every element of G must converge strongly on every element of the algebra. The proof works by turning an arbitrary completely positive map into a weak limit of representations via Stinespring dilation and a unilateral shift, so that the representation-only convergence condition forces the original map to be a representation. The result completes the equivalence that makes hyperrigidity a purely representation-theoretic notion in the separable case.","feed_headline":"Hyperrigidity reduces to representation limits alone","feed_subtitle":"Now hyperrigidity can be tested by representations alone, no completely positive maps required.","key_machinery":"The carrying object is a unilateral shift of multiplicity ℵ0 used to build approximations to a completely positive map. Given a UCP map Φ = Pρ(·)|_H from Stinespring dilation, the proof forms the isometry V = I_H ⊕ S on K = H ⊕ (K⊖H), where S shifts the complementary summand, and defines π_m(a) = V^m ρ(a) $V^{{*m}}$ + (I - V^m $V^{{*m}}$)σ∞(a), with σ∞ the infinite direct sum of the tail representation σ. As m tends to infinity, $V^{{*m}}$ tends strongly to the projection onto H, so the weak limit of these representations is Pρ(·)P ⊕ σ ⊕ σ ⊕ ..., which agrees with Φ on G. Because V and $V^{{*m}}$ act diagonally against σ∞, each π_m is a genuine representation, not merely a positive map. The shift is what converts a completely positive map into a weak limit of representations.","core_discovery":"The central claim is Theorem 2.1: if A is a unital separable C*-algebra and G is nonempty, then G is hyperrigid if and only if for every Hilbert space H, every representation π: A → B(H), and every sequence of representations π_n: A → B(H), weak-operator convergence of π_n(g) to π(g) for all g ∈ G implies strong-operator convergence of π_n(a) to π(a) for all a ∈ A. The forward direction is inherited from known characterizations of hyperrigidity. The reverse direction is the new contribution: starting from a unital completely positive map Φ that agrees with a representation π on G, Stinespring dilation represents Φ as a compression of a representation ρ, and an amplification plus shift construction produces genuine representations whose weak-operator limit is Φ ⊕ σ ⊕ σ ⊕ ..., so the assumed representation-only property forces π = Φ. Separability enters by guaranteeing that the Hilbert spaces can be taken separable and that an infinite-dimensional separable representation is available to serve as the tail in the construction.","pith_inferences":["Editorial inference: a natural testable extension is to seek a proof that avoids the cardinality reduction, which would extend the equivalence to nonseparable C*-algebras; the current proof depends on separability precisely at that reduction.","Editorial inference: the representation-only reformulation may give a concrete route toward Arveson's hyperrigidity conjecture in the commutative case, where checking whether any pair of representation sequences splits becomes a potentially computable criterion.","Editorial inference: in approximation practice, the theorem suggests a finite-dimensional testing scheme: approximate candidate generators by finite-dimensional representations and look for weak convergence that fails to be strong.","Editorial inference: the proof's shift construction is likely adaptable to other settings where a positive map is given by a compression of a representation, such as semispectral measures and moment problems, wherever an infinite-dimensional separable tail representation exists."],"forward_implications":["Hyperrigidity of a set G in a separable unital C*-algebra is equivalent to a statement about *-homomorphisms alone: any sequence of representations that converges weakly on G must converge strongly everywhere.","To certify hyperrigidity, one no longer needs to reason about unital completely positive maps; it is enough to show that representation sequences cannot split in the weak-to-strong sense.","The unique extension property, the original hyperrigidity definition, and the representation-only convergence condition all coincide for separable algebras.","To disprove hyperrigidity of G, it suffices to exhibit a representation π and representations π_n with π_n(g) → π(g) weakly for all g ∈ G but π_n(a) ↛ π(a) strongly for some a ∈ A."],"supporting_citations":[{"why":"Arveson's paper supplies the definition of hyperrigidity and the characterization through the unique extension property used as Theorem 1.2(i)-(iii).","marker":"[3]"},{"why":"Kleski's result provides the weak-operator-topology characterization of hyperrigidity for UCP maps, which is condition (iv) in Theorem 1.2.","marker":"[20]"},{"why":"The authors' companion paper supplies the separability-dependent version of Theorem 1.2 that connects conditions (i)-(iv), and it is cited as Theorem B.2.","marker":"[22]"},{"why":"Stinespring's dilation theorem provides the representation ρ whose compression gives the UCP map Φ, the starting point of the shift construction.","marker":"[28]"}],"fun_headline_variants":["Hyperrigidity via representations only","Representations suffice for hyperrigidity","Hyperrigidity without completely positive maps","Representation limits decide hyperrigidity"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument that a purely representation-theoretic convergence condition forces hyperrigidity requires the ambient C*-algebra A to be separable; without separability the proof cannot restrict to separable Hilbert spaces or build the infinite tail representation used in the shift construction.","fun_headline_variants_meta":{"raw":{"variants":["Hyperrigidity via representations only","Representations suffice for hyperrigidity","Hyperrigidity without completely positive maps","Representation limits decide hyperrigidity"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000663,"raw_usage":{"total_tokens":2936,"prompt_tokens":758,"completion_tokens":2178,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":374,"completion_tokens_details":{"reasoning_tokens":2128}},"tokens_in":374,"tokens_out":2178,"duration_ms":15849,"temperature":1.0,"reasoning_tokens":2128,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T04:18:29.563283+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find a unital C*-algebra A and a nonempty subset G such that every sequence of representations that converges weakly on G converges strongly on A, yet some sequence of unital completely positive maps converges weakly on G without converging strongly on all of A; in the separable case the theorem predicts such a pair cannot exist, so any concrete example would refute the characterization.","supporting_citations":[{"cited_title":"Arveson, The noncommutative Choquet boundary II: hyp errigidity, Israel J","cited_arxiv_id":null,"evidence_quote":"Arveson's paper supplies the definition of hyperrigidity and the characterization through the unique extension property used as Theorem 1.2(i)-(iii)."},{"cited_title":"Kleski, Korovkin-type properties for completely po sitive maps, Illinois J","cited_arxiv_id":null,"evidence_quote":"Kleski's result provides the weak-operator-topology characterization of hyperrigidity for UCP maps, which is condition (iv) in Theorem 1.2."},{"cited_title":"Pietrzycki, J","cited_arxiv_id":null,"evidence_quote":"The authors' companion paper supplies the separability-dependent version of Theorem 1.2 that connects conditions (i)-(iv), and it is cited as Theorem B.2."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Stinespring's dilation theorem provides the representation ρ whose compression gives the UCP map Φ, the starting point of the shift construction."}],"review_version":1}