{"id":"fbccdbab-7ade-4928-90c0-eea7b10ee3da","arxiv_id":"2412.15733","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"An exotic locally compact group with a Borel 2-cocycle is built where the untwisted group von Neumann algebra is a factor and the twisted algebra has a diffuse center.","lead":"This paper constructs a locally compact group whose untwisted group von Neumann algebra is a factor, but whose 2-cocycle twisted version has a diffuse center. The example answers a natural question about braided tensor products of von Neumann algebras and rules out an intrinsic factoriality criterion in the locally compact case.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 4's essential-freeness proof assumes R has no zero divisors, but the ring R=∏'(Q_p,Z_p) in Lemma 5(iii) has zero divisors, so the proof of factoriality of L(G) is incomplete as written.","rationale":"The reader's weakest_assumption correctly identifies the zero-divisor gap in Lemma 4 as applied to Lemma 5(iii). This is the single most load-bearing concern because it blocks the proof that L(G) is a factor in the main example. The concern is a genuine proof gap, not a demonstrated false theorem: the intended finite-support structure of Γ should repair it, but that repair is not supplied in the manuscript. I checked the surrounding arguments for other potential weaknesses. The ergodicity proof in Lemma 5(iii) is plausible and does not depend on zero divisors. The construction of the twisted algebra with diffuse center ultimately relies on Z(Lω(G))≅Z(L(Γ)) rather than on the factoriality of L(G), so the diffuse-center conclusion is not threatened by the freeness gap; the missing piece is specifically the factoriality of L(G). The paper contains some notational overloading in Lemma 3, but the intended relation Z(Lω(G))≅Z(L(Γ)) is recovered in Lemma 4's statement and proof, so I do not treat that as the load-bearing issue. Because the identified concern exactly matches the reader's weakest assumption and does not move the overall conditional assessment, the verdict should remain unchanged.","tokens_in":6907,"tokens_out":25674,"duration_ms":231290,"concrete_test":"Prove essential freeness directly for Lemma 5(iii): fix a nontrivial finite-support sequence A=(A_k)∈SL2(Q)^(N), choose k with A_k≠I_2, and show that the set {x∈R^2 : A_k x_k = x_k for all k} has Haar measure zero by observing that its k-th coordinate lies in a line of Q_p^2 and applying the cylinder definition of the restricted product measure. If this argument succeeds and no other use of the zero-divisor hypothesis remains, the gap is repaired and the central claim stands; if it fails for some A, compute the measure of the fixed set explicitly to see whether ergodicity alone still forces L(G) to be a factor.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central construction needs L(G) to be a factor, and Lemma 4 derives this from essential freeness of Γ↷R^2. Its proof argues that for a nonzero row (a,b), the set D={(x,x') : ax+bx'=0} has measure zero 'because R has no zero divisors, there is at most one x' for each x'. This hypothesis is false for R=∏'_{k∈N}(Q_p,Z_p) used in Lemma 5(iii): elements supported on disjoint coordinates multiply to zero. For example, a=(0,1,0,0,...) and b=(1,0,0,...) give ab=0, and for x'=0, every x with x_2=0 satisfies ax=0, so the fiber is not a singleton. Thus the measure-zero argument via Fubini in Lemma 4 does not apply, and the statement in Lemma 5(iii) that all assumptions of Lemma 4 are satisfied is not correct as written. The gap is likely repairable: Γ=SL2(Q)^(N) is a finite-support subgroup, so for any nontrivial A there is a coordinate k with A_k≠I_2; in that coordinate the fixed-point set is a proper line in Q_p^2, which has Haar measure zero, and the full fixed-point set is contained in the corresponding null cylinder. But this finite-support argument is absent, so the proof of factoriality of L(G) in Theorem A is incomplete as written, even though the theorem itself may well be true.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper constructs, for each odd prime p, a locally compact second countable group G (a restricted product of semidirect products Q_p^2 ⋊ SL_2(Q)) and a Borel 2-cocycle ω such that the untwisted group von Neumann algebra L(G) is a factor while the twisted algebra L_ω(G) has a diffuse center. The construction proceeds through a sequence of lemmas: a Takesaki-duality-based crossed-product identity, a realization of inner crossed products as twisted group algebras, a reduction of the center of the twisted algebra to the center of an ergodic crossed product, and an ergodic-theoretic lemma applied to restricted products of p-adic fields. The paper also proves that for discrete G, factoriality of L(G) and A forces A ⋊ G to be a factor, and that this permanence fails without discreteness.","tokens_in":7206,"tokens_out":22862,"duration_ms":196895,"significance":"If the proof is completed, Theorem A provides a striking dichotomy between twisted and untwisted group von Neumann algebras in the locally compact setting, supporting the paper's claim that no intrinsic group-theoretic criterion can characterize factoriality for twisted algebras in that generality. The argument is constructive and uses standard tools (Takesaki duality, Moore ergodicity, Borel selection) rather than fitted parameters, and the counterexample to the discrete permanence statement in Proposition B is relevant for applications in braided tensor products. The theorem itself is very plausible and the gaps identified below are local and repairable, but the written proof is currently incomplete.","major_comments":[{"comment":"Lemma 5(iii) states that R = ∏'(Q_p, Z_p) satisfies all assumptions of Lemma 4. This is not correct: R has zero divisors. For example, a = (1, 0, 1, 0, ...) and b = (0, 1, 0, 1, ...) are nonzero but ab = 0. The proof of essential freeness in Lemma 4 relies on the absence of zero divisors to conclude that for fixed x' the equation ax + bx' = 0 has at most one solution; for this ring that inference fails. The gap is repairable by a finite-support argument: for any A ≠ I_2 in SL_2(R), choose a coordinate k with A_k ≠ I_2; the fixed-point set in that coordinate is a proper subspace of Q_p^2, hence of Haar measure zero, and the full fixed-point set is contained in the corresponding null cylinder. But as written, the deduction of Theorem A from Lemmas 4 and 5(iii) is incomplete.","section":"§2, Lemma 5(iii)"},{"comment":"The displayed bicharacter identity in Lemma 3, Ω(θ(y,φ), θ(y′,φ′)) = φ′(y) φ(y′), is symmetric under interchange of the two arguments. If it held literally, L_Ω(S) would be commutative and could not be isomorphic to B(L^2(T)) for nontrivial T. The computation in the proof of Lemma 3 (λ_Ω(θ(y,1)) λ_Ω(θ(0,φ)) = φ(y)^2 λ_Ω(θ(0,φ)) λ_Ω(θ(y,1))) requires the antisymmetric form Ω(θ(y,φ), θ(y′,φ′)) = φ′(y) \\overline{φ(y′)} (equivalently φ′(y) φ(y′)^{-1}); the printed formula appears to be missing the conjugate/inverse. This should be corrected. With the corrected formula, the proof of Lemma 3 is consistent, and Lemma 4's symplectic bicharacter ψ(xy′ − x′y) is of the required form.","section":"§2, Lemma 3"}],"minor_comments":[{"comment":"Lemma 5(ii) states that Z(L(Γ)) is 2n-dimensional for Γ = SL_2(Z)^n or SL_2(Q)^n. The center of each factor is {±I_2}, so the center of Γ is {±I_2}^n, a group of order 2^n. The corresponding von Neumann algebra has 2^n minimal projections, hence is 2^n-dimensional as a vector space, not 2n-dimensional. This does not affect Theorem A but should be corrected.","section":"§2, Lemma 5(ii)"},{"comment":"Lemma 4's proof uses the fact that x ↦ 2x is a homeomorphism of R, but the lemma statement does not include invertibility of 2 in R. Add this as an explicit hypothesis (or state the equivalent condition used to apply Lemma 3). All examples in Lemma 5 satisfy it, but the lemma as stated is broader than its proof supports.","section":"§2, Lemma 4"},{"comment":"In the last line of the proof of Proposition B, 'b ∈ C1' should be 'b ∈ C·1'; this is a typographical issue only.","section":"§3, Proposition B"}],"recommendation":"major_revision","confidential_remarks":"Both major comments are local and repairable. The zero-divisor gap can be closed by a finite-support coordinate argument, and the Lemma 3 bicharacter formula is clearly missing a conjugate or inverse. I do not see grounds to doubt the truth of Theorem A, and I recommend asking for a revised version rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The headline result is real and worth knowing: Vaes constructs a locally compact group G such that L(G) is a factor but a 2-cocycle twist L_ω(G) has diffuse center, killing any intrinsic group-theoretic criterion for twisted factoriality in the locally compact case. That is a substantial counterexample, and it resolves the DCK24 question about braided tensor products of factors. The construction is clever: a restricted product of Q_p^2 ⋊ SL2(Q), a bicharacter twist, and Takesaki duality to convert the twisted algebra into a crossed product of B(K). Lemma 3 is the technical core and it is genuinely elegant. Proposition B is a nice complement, and the discrete case proof is clean.\n\nThe soft spot is exactly where the reader says. Lemma 4 requires R to have no zero divisors to prove essential freeness of Γ ↷ R^2, using the claim that ax = 0 has at most one solution. Lemma 5(iii) then asserts this applies to R = ∏'(Q_p, Z_p). But that ring has zero divisors: an element supported on coordinate 1 times one supported on coordinate 2 is zero. So the proof of factoriality of L(G) in Theorem A is incomplete as written. This is a genuine gap, not a stylistic quibble.\n\nThat said, the gap is small and almost certainly repairable. Γ = SL2(Q)^{(N)} is finite-support, so any non-identity matrix is non-identity in some coordinate k, and its fixed-point set in Q_p^2 is a proper line of Haar measure zero; the full fixed set is then contained in a null cylinder. A short finite-support argument would patch Lemma 5(iii). The paper does not include it, so the theorem remains conditional until it is added.\n\nThe citation pattern is clean: Kleppner and DCK24 are cited as motivation and background, not to define the result. This is not a case of self-citation inflation. The paper is otherwise well argued and the main construction is convincing modulo the gap.\n\nWho gets value: operator algebraists working on group von Neumann algebras, crossed products, or braided tensor products. It deserves serious referee time. Recommendation: send to peer review, and ask the author to repair Lemma 4/5(iii) with a finite-support argument. I would not desk reject.","headline":"A genuinely new counterexample in twisted group von Neumann algebras, with a real but likely repairable gap in the proof of Lemma 4 that currently leaves Theorem A conditional.","tokens_in":7736,"tokens_out":2958,"would_cite":true,"duration_ms":26207,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["46L10","46L55","22D25","22E40"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper constructs a locally compact group whose untwisted group von Neumann algebra is a factor while a 2-cocycle twist of the same algebra has a diffuse center, and proves the discrete case cannot exhibit this behavior.","keywords":["group von Neumann algebra","factor","twisted group von Neumann algebra","Borel 2-cocycle","locally compact group","restricted product","p-adic field","diffuse center"],"falsifier":"Check whether $\\mathrm{SL}_2(\\mathbb{Q})^{(\\mathbb{N})}$ acts essentially freely on $R^2$ when $R$ is the restricted product $\\prod'_{k\\in\\mathbb{N}}(\\mathbb{Q}_p,\\mathbb{Z}_p)$: if some nontrivial element has a fixed set of positive measure, Lemma 4's conclusion collapses, while a direct finite-support proof of essential freeness would repair the gap.","tokens_in":1664,"feed_emoji":"🌀","tokens_out":2888,"duration_ms":114716,"temperature":0.7,"pith_summary":"For discrete groups, a group von Neumann algebra is a factor exactly when every nontrivial conjugacy class is infinite, and then every 2-cocycle twist is still a factor. This paper shows the locally compact case is different: it constructs a locally compact group and a Borel 2-cocycle such that the untwisted group von Neumann algebra is a factor, while the twisted one has a diffuse center. The construction uses a restricted product of p-adic semidirect products and transfers the center of an auxiliary group algebra into the twisted algebra of a new group, while making the new group's untwisted algebra a factor. A companion result shows this phenomenon cannot happen for discrete groups, where factoriality of the group algebra and of the acting algebra always forces factoriality of the crossed product. If correct, the example rules out any intrinsic group-theoretic criterion for factoriality of twisted locally compact group von Neumann algebras.","feed_headline":"Twisting breaks factoriality for a locally compact group","feed_subtitle":"A factor untwisted, diffuse after a twist: no group-theoretic test can determine factoriality.","key_machinery":"The engine is a transfer principle for centers. Given an action $G \\curvearrowright S$ of a lcsc group on a lcsc abelian group and a $G$-invariant alternating bicharacter $\\Omega: S \\times S \\to \\mathbb{T}$ satisfying a self-duality condition $S \\cong T \\times \\widehat{T}$, the semidirect product $\\mathcal{G} = \\widehat{S} \\rtimes_{\\widehat{\\alpha}} G$ carries a Borel 2-cocycle $\\omega$ whose twisted center is isomorphic to the untwisted center of $L(\\mathcal{G})$, while $L(\\mathcal{G})$ is stably isomorphic to $L^\\infty(S) \\rtimes_\\alpha G$. In the application, $S = R^2$, $G = \\Gamma$, and $\\Omega((x,x'),(y,y')) = \\psi(xy' - x'y)$, with $R$ a restricted product of p-adic fields and $\\psi$ a character summing residue classes. The roles are split: ergodicity of $\\Gamma$ on $R^2$ makes $L(\\mathcal{G})$ a factor, while an infinite central subgroup of $\\Gamma$ makes the transferred center diffuse.","core_discovery":"Theorem A states the main example: for an odd prime p, take the locally compact group $G = \\mathbb{Q}_p^2 \\rtimes \\mathrm{SL}_2(\\mathbb{Q})$ with compact open subgroup $K = \\mathbb{Z}_p^2$, and form the restricted product $\\mathcal{G} = \\prod'_{k\\in\\mathbb{N}}(G,K)$. Then $\\mathcal{G}$ carries a Borel 2-cocycle $\\omega$ such that $L(\\mathcal{G})$ is a factor while $L_\\omega(\\mathcal{G})$ has a diffuse center. The proof passes through a series of lemmas: a dual semidirect product construction shows that the twisted center of the new group is isomorphic to the untwisted center of an auxiliary group, while the untwisted algebra of the new group is, up to stabilization, a crossed product by an ergodic action. Taking the auxiliary group to be $\\Gamma = \\mathrm{SL}_2(\\mathbb{Q})^{(\\mathbb{N})}$ acting on $R^2$, where $R$ is the restricted product of p-adic fields with p-adic integers, the action is ergodic, so $L(\\mathcal{G})$ is a factor; but $\\Gamma$ contains the infinite central subgroup $\\{\\pm I\\}^{(\\mathbb{N})}$, so the center of $L(\\Gamma)$, hence of the twisted algebra, is diffuse. Proposition B completes the picture by showing that for discrete $G$, factoriality of $L(G)$ and of $A$ always implies factoriality of $A \\rtimes_\\alpha G$.","pith_inferences":["Extending beyond the paper, the same transfer principle could be iterated to encode the center of an arbitrary discrete group algebra into the twisted center of a locally compact group algebra, potentially realizing many prescribed centers as $Z(L_\\omega(\\mathcal{G}))$.","The construction likely works for any countable subgroup $\\Gamma$ of $\\mathrm{SL}_2(R)$ with an ergodic action on $R^2$; the center of the twisted algebra will then be isomorphic to $Z(L(\\Gamma))$, giving a zoo of factoriality behaviors controlled by central subgroups.","For quantum-group braided tensor products, the example implies that the braiding can destroy factoriality even when both factors are type I factors, which may be relevant beyond the crossed-product reformulation used in the paper."],"forward_implications":["If Theorem A is correct, no group-theoretic invariant computed from $\\mathcal{G}$ alone can decide when $L_\\omega(\\mathcal{G})$ is a factor: the same underlying group is factorial for $\\omega = 1$ and non-factorial for a suitable $\\omega$.","The example gives a counterexample to the natural expectation that braided tensor products of factors are factors, producing an action of a locally compact group on $B(K)$ with both $L(\\mathcal{G})$ and $B(K)$ factors but the crossed product not a factor.","The discrete/locally compact boundary is sharp: for discrete $G$, factoriality of $L(G)$ and $A$ implies factoriality of $A \\rtimes_\\alpha G$, so the counterexample necessarily uses a nondiscrete group.","Restricted products of p-adic semidirect products provide a flexible source of such factoriality-splitting groups, and varying the auxiliary group $\\Gamma$ changes the center of the twisted algebra in a controlled way."],"supporting_citations":[{"why":"Supplies the Borel selection theorem used in Lemma 2 to lift an inner action to a Borel unitary cocycle, converting the crossed product into a twisted group von Neumann algebra.","marker":"[Sri80]"},{"why":"Provides the known criterion for factoriality of twisted group von Neumann algebras of discrete groups, the classical contrast that makes the locally compact example exotic.","marker":"[Kle61]"},{"why":"Raises the question whether braided tensor products of factors are factors and records the application, Corollary 8.5, where Theorem A is used to build a counterexample.","marker":"[DCK24]"}],"fun_headline_variants":["A twist that turns a factor into an algebra with diffuse center","For a certain group, a twist yields a diffuse center from a factor","Twisted group algebras: factor to diffuse center","Locally compact group: twist gives diffuse center, not factor","2-cocycle turns a group factor into diffuse-center algebra"],"cache_read_input_tokens":9856,"weakest_assumption_plain":"The load-bearing premise is Lemma 4's assumption that the ring $R$ has no zero divisors, which the proof uses to show the action $\\Gamma \\curvearrowright R^2$ is essentially free; the ring built from a restricted product of p-adic fields does have zero divisors, so the printed proof does not cover the main example.","fun_headline_variants_meta":{"raw":{"variants":["A twist that turns a factor into an algebra with diffuse center","For a certain group, a twist yields a diffuse center from a factor","Twisted group algebras: factor to diffuse center","Locally compact group: twist gives diffuse center, not factor","2-cocycle turns a group factor into diffuse-center algebra"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000563,"raw_usage":{"total_tokens":2651,"prompt_tokens":901,"completion_tokens":1750,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":517,"completion_tokens_details":{"reasoning_tokens":1665}},"tokens_in":517,"tokens_out":1750,"duration_ms":15136,"temperature":1.0,"reasoning_tokens":1665,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T11:09:42.507959+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Check whether $\\mathrm{SL}_2(\\mathbb{Q})^{(\\mathbb{N})}$ acts essentially freely on $R^2$ when $R$ is the restricted product $\\prod'_{k\\in\\mathbb{N}}(\\mathbb{Q}_p,\\mathbb{Z}_p)$: if some nontrivial element has a fixed set of positive measure, Lemma 4's conclusion collapses, while a direct finite-support proof of essential freeness would repair the gap.","supporting_citations":[],"review_version":1}