{"id":"6081caa5-d65d-4681-9004-4d7573147dcb","arxiv_id":"2608.12080","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"For every d≥2, the canonical word-length spectral triple of (Z/2Z)≀F_d is not a spectral metric space, giving the first family of such counterexamples.","lead":"This paper proves that the lamplighter groups built from the free group on d generators, for every d at least 2, produce word-length spectral triples that are not spectral metric spaces. It supplies the first explicit family of counterexamples to a question in non-commutative metric geometry that had been expected but never answered.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified; the proof's most exposed estimate, Proposition 2.3, withstands close review.","rationale":"The reader's ACCEPT verdict is well supported. I conducted an independent close reading of the full argument, with special attention to Proposition 2.3 and the operator norm estimate in Proposition 3.3, the two places where an error would be most damaging. Proposition 2.3 is the genuinely exposed step: it has to hold for every E,g,x and every finite symmetric generating set S. The proof is nevertheless sound. Lemma 2.1 provides the geometric localization that makes both directions of the word-length comparison work; the constants R_S, J_S, c_S, d_S are finite because S is finite and generates G. The reverse inequality (2.2) is even uniform in the separation r, which is what makes the global bound possible. The transfer to commutators in Theorem 3.4 is exact, including the x in E case via H_x(E△{x},g)=H_x(E,g). Proposition 3.3's bound on B_r is a clean orthogonal-decomposition argument; the geometric series converges precisely because d≥2. The separation of the z_r via the character χ on the amenable lamp subgroup is standard. The GPT-5.6 Sol acknowledgment is a transparency statement and does not interact with the mathematics. No significant objection is identified.","tokens_in":10006,"tokens_out":24078,"duration_ms":222827,"concrete_test":"Implement Proposition 2.3 for the standard generating set of (Z/2Z)≀F_2 (one toggle at the identity and the four base moves), enumerate all E,g,x with d_T(e,g)≤4 and #E≤4, and verify |ℓ_S(χ_{E∪{x}},g)-ℓ_S(χ_E,g)| ≤ C_S(1+d_T(x,H(E,g))) with C_S computed as in the proof; repeat for one non-standard generating set to confirm the constant remains finite.","verdict_should_be":"UNCHANGED","load_bearing_attack":"I read the proof as a chain: Proposition 2.3 converts tree distance into word-length variation; Theorem 3.4 turns this into a uniform commutator bound via the operator B_r whose norm is controlled in Proposition 3.3; the separated sequence is then produced by a character on the lamp subgroup. The weakest step is Proposition 2.3, exactly as the reader noted. I checked its proof in detail. The upper bound (2.1) uses Lemma 2.1 to place the new lamp x within R_S+J_S of a prefix base point, so inserting the toggle costs at most 2c_S(r+R_S+J_S)+d_S; the lower bound (2.2) is even cheaper because x is already within R_S of some prefix in a word for the enlarged configuration. The constant C_S = 2c_S(R_S+J_S+1)+d_S therefore dominates both. The application to Theorem 3.4 is legitimate: for x in E one applies Proposition 2.3 to E\\{x}, and H_x(E△{x},g)=H_x(E,g) makes the pointwise inequality exact. Proposition 3.3's norm bound for B_r is also sound: the Q_{x,r} operators decompose ℓ^2(G) into two orthogonal pieces and have norm sqrt(#(S_r∩T_x)). I found no hidden dependence on S beyond the finite C_S and no circularity.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the canonical word-length spectral triple (C[G], ℓ²(G), D_ℓ) associated with a countable discrete group G and a proper length function, asking whether the Connes pseudo-metric metrizes the weak-∗ topology on the state space of the reduced group C∗-algebra, i.e., whether the triple is a compact quantum metric space in Rieffel's sense. The main result, Theorem A, asserts that for the lamplighter group G = (Z/2Z)≀F_d with d ≥ 2, equipped with the word-length ℓ_S coming from any finite symmetric generating set S, this spectral triple is not metric. The proof constructs elements z_r that average the lamp toggles over the sphere S_r of radius r in the Cayley tree of F_d, proves via a local perturbation inequality (Proposition 2.3) and an operator-norm estimate (Proposition 3.3) that the commutator seminorm of z_r is uniformly bounded in r, and then shows that the normalized elements form a uniformly separated sequence in the set {a ∈ C[G] : τ(a) = 0, ||[D_ℓ,a]|| ≤ 1}. By the Ozawa–Rieffel characterization (Proposition 1.3), total boundedness of this set is necessary for a compact quantum metric space, yielding the contradiction.","tokens_in":10254,"tokens_out":20081,"duration_ms":172723,"significance":"If correct, Theorem A provides the first explicit family of groups for which the canonical word-length spectral triple fails to be a compact quantum metric space, answering a question that the paper documents as open in the literature (see [2]). The proof is explicit and self-contained: Proposition 2.3 tracks all constants in terms of the generating set, Proposition 3.3 reduces the operator norm to a geometric-series computation, and the final separation uses a concrete character of the lamp subgroup. There are no fitted parameters, no hidden normalization choices, and no reliance on the author's earlier work. The result is likely to stimulate further investigation of generating-set independence for metric spectral triples and of wreath products over hyperbolic groups (Question 3.5). The paper is a clean, publishable contribution to noncommutative metric geometry.","major_comments":[],"minor_comments":[{"comment":"The proof says 'We may assume m ≥ 1' but does not explicitly handle the degenerate case m = 0, i.e., (χ_E,g) = e. In that case the claimed inequality reduces to ℓ_S(χ_{{x}},e) ≤ C_S(1+d_T(x,e)), which follows from the same estimates; please add a sentence covering this case.","section":"Section 2, Proposition 2.3"},{"comment":"When defining L1 and asserting that \tilde z_r belongs to L1, the condition τ(\tilde z_r) = 0 is not explicitly verified. It holds because each unitary λ_{(χ_{{g}},e)} with g ∈ S_r and r ≥ 1 has trivial trace; please spell this out for completeness.","section":"Section 3, Theorem A"},{"comment":"The identity B_r = (1/#S_r)Σ_{x∈S_r}λ(χ_{{x}},e) + (1/#S_r)Σ_{k=1}^r Σ_{x∈S_k} Q_{x,r} is stated without derivation. A short explanation that the coefficient of δ_{(χ_{E△{y}},g)} in the double sum equals Σ_{k=1}^r δ(g∉T_{v_k(y)})δ((E\\{y})∩T_{v_k(y)}=∅), which by Lemma 3.1 coincides with d_T(y,H_y(E,g)), would improve readability.","section":"Section 3, Proposition 3.3"},{"comment":"After combining (2.1) and (2.2), the text does not explicitly note that 2c_S R_S + d_S is bounded by C_S and that 2c_S(r+R_S+J_S)+d_S is bounded by C_S(1+d_T(x,H(E,g))); adding this one-line estimate would make the final constant C_S fully transparent.","section":"Section 2, Proposition 2.3"}],"recommendation":"minor_revision","confidential_remarks":"This is a solid, publishable note. The central claim is sound and the proof is explicit, with all constants tracked. The minor comments are presentation-level and can be addressed without changing the mathematics. The editor may wish to verify the literature claim that no counterexample was previously known, but this does not affect the technical correctness of the paper."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nShort version: this note delivers what it claims. For every d ≥ 2 and every finite symmetric generating set of the lamplighter group (Z/2Z)≀F_d, the canonical word-length spectral triple is not a spectral metric space. That is the first explicit family of groups for which a word-length spectral triple fails to be a compact quantum metric space, resolving an open problem that Austad's survey [2] explicitly flags as having no known example. I checked the proof chain and it holds.\n\nWhat is new: the construction. The paper defines averaging operators z_r over the sphere of radius r in the tree, uses a local perturbation estimate (Proposition 2.3) to control the commutator [D, z_r] uniformly in r, and then separates the z_r via a character on the abelian lamp subgroup. The technical vehicle is not a routine transfer of earlier positive results; it is tailored to the wreath product structure. The proof is self-contained after the standard Ozawa–Rieffel characterization of Lip-norms.\n\nThe weakest step, as the reader also notes, is Proposition 2.3, which bounds word-length variation when toggling a lamp at distance r from the hull H(E,g). I went through it carefully. The upper bound (2.1) uses Lemma 2.1 to place the new lamp within R_S + J_S of a prefix base point; the lower bound (2.2) is even cheaper. The constant C_S = 2c_S(R_S+J_S+1)+d_S dominates both, and the application in Theorem 3.4 is exact because H_x(E△{x},g)=H_x(E,g). Proposition 3.3's norm bound for B_r via the Q_{x,r} is also sound; the decomposition into orthogonal pieces is clean.\n\nSoft spots are minor. The paper is a short note, so it leaves broader questions open—e.g., whether the metric property depends on the generating set. That is not a flaw. The acknowledgement of GPT-5.6 Sol as an exploratory tool is transparent and does not affect the mathematics; the arguments are explicitly verified by the author. Citation pattern looks appropriate: the key external input is Ozawa–Rieffel, and the survey [2] is cited for the open problem. No fitted parameters, no hidden assumptions.\n\nFor whom: anyone working in non-commutative metric geometry or coarse geometry of groups. It is a clean counterexample that should be cited. I would send it to a serious referee. Given that the central proof is explicit and complete, my recommendation is accept after a routine check; nothing here warrants heavy revision.","headline":"First explicit counterexample to metric word-length spectral triples; proof is sound and self-contained, worth refereeing.","tokens_in":10765,"tokens_out":2643,"would_cite":true,"duration_ms":23638,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["46L87","58B34","46L89","20F65"],"pacs":[],"model":"deepseek-v4-flash","headline":"For every integer d ≥ 2 and every finite symmetric generating set, the word-length spectral triple of the lamplighter group (Z/2Z)≀F_d is not a spectral metric space.","keywords":["Lamplighter group","spectral triple","compact quantum metric space","word-length function","free group","noncommutative metric geometry","Lip-norm","state space"],"falsifier":"Take $d=2$ with the standard generators of the free group and compute the operator norm of $[D_\\ell,z_r]$ for growing $r$: Theorem A predicts a uniform upper bound, so any unbounded growth would disprove the main claim. Separately, one can hunt for a configuration $E,g,x\\notin E$ where the left-hand side of Proposition 2.3 exceeds any constant multiple of $1+d_T(x,H(E,g))$; such a configuration would directly falsify the key estimate on which the proof rests.","tokens_in":9788,"feed_emoji":"💡","tokens_out":13065,"duration_ms":120549,"temperature":0.7,"pith_summary":"This paper proves the first explicit family of counterexamples to a long-standing expectation in noncommutative metric geometry: word-length spectral triples need not be metric. For every $d\\ge 2$, take the lamplighter group $G=(\\mathbb{Z}/2\\mathbb{Z})\\wr\\mathbb{F}_d$ and any finite symmetric generating set $S$; the diagonal operator $D_\\ell$ on $\\ell^2(G)$ whose eigenvalues are word lengths yields a spectral triple $(C[G],\\ell^2(G),D_\\ell)$. The theorem states that this triple is never a spectral metric space: the pseudometric it induces on the state space of the reduced group $C^*$-algebra does not recover the weak-$^*$ topology. The result matters because it closes a question that was explicitly open, showing that compact quantum metric spaces fail for a concrete, natural family of groups rather than only in hypothetical constructions.","feed_headline":"First counterexample: lamplighter word-length triples are not metric","feed_subtitle":"For every d≥2 and every generating set, the diagonal operator on the lamplighter group fails to metrize the state space.","key_machinery":"The load-bearing construction is the family $z_r=\\frac{1}{\\#S_r}\\sum_{g\\in S_r}\\lambda(\\chi_{\\{g\\}},e)$, the average of lamp toggles at all points of the radius-$r$ sphere in the Cayley tree of $\\mathbb{F}_d$. Its bounded commutator rests on Proposition 2.3, a local perturbation estimate: adding a lamp at a point $x$ to a configuration $E$ changes word length by at most $C_S(1+d_T(x,H(E,g)))$, where $H(E,g)$ is the smallest subtree containing $E$, $e$, and $g$, and $C_S$ depends only on the generating set. This estimate lets the commutator $[D_\\ell,z_r]$ be dominated by sums of operators $Q_{x,r}$ whose norms decay geometrically in the tree distance, giving a uniform bound independent of $r$. Separation comes from the abelian structure of the lamp subgroup: a multiplicative character on the lamp coordinates distinguishes $z_r$ and $z_{r'}$ by a fixed amount. Finally a total-boundedness criterion for compact quantum metric spaces converts the separated sequence into a proof that the triple is not metric.","core_discovery":"The paper's central claim is Theorem A: for every integer $d \\ge 2$, if $G=(\\mathbb{Z}/2\\mathbb{Z})\\wr\\mathbb{F}_d$ and $S$ is any finite symmetric generating set with word-length function $\\ell=\\ell_S$, then the spectral triple $(C[G],\\ell^2(G),D_\\ell)$ is not a spectral metric space. The proof produces, for each radius $r$, an element $z_r$ in the group algebra obtained by averaging the lamp-toggling unitaries over the sphere of radius $r$ in the Cayley tree of $\\mathbb{F}_d$. These elements have uniformly bounded commutator with $D_\\ell$, so a suitable rescaling puts them all in the unit Lipschitz ball; yet they are mutually separated by a positive constant in norm, detected through a character (a multiplicative state) of the abelian lamp subgroup. Since a compact quantum metric space would force the unit Lipschitz ball to be totally bounded, the infinite separated sequence shows the metric property fails. The argument works uniformly in the generating set, so the counterexample is not tied to a special choice of word-length.","pith_inferences":["The same averaging-and-separation strategy may generalize to wreath products $\\mathbb{Z}/2\\mathbb{Z}\\wr\\Gamma$ with $\\Gamma$ any non-elementary hyperbolic group, using hyperbolic geodesics in place of the tree; the paper states this as an open question rather than a claim.","Because the separation is detected by a single character of the abelian lamp subgroup, replacing the lamp group by any finite abelian group with a nontrivial character is a plausible route to further counterexamples; this is an extension the paper does not pursue.","The local perturbation estimate Proposition 2.3 is the natural place to probe the dividing line: if a group admits a length function with a finite-range-to-tree-distance bound and a spheres-indexed separated sequence, metricity fails; searching for such bounds may yield a criterion for metric spectral triples."],"forward_implications":["For every finite symmetric generating set of $(\\mathbb{Z}/2\\mathbb{Z})\\wr\\mathbb{F}_d$, $d\\ge2$, the word-length spectral triple fails to be a compact quantum metric space; the obstruction is intrinsic to the group and the word-length construction, not to one generating set.","The positive results known for polynomial-growth groups and word-hyperbolic groups cannot be extended to all finitely generated groups; the lamplighter groups form a concrete boundary case.","The unit Lipschitz ball contains an infinite, uniformly separated sequence, so the pseudometric induced on the state space cannot coincide with the weak-$^*$ topology for these triples.","This is the first explicit family of groups settling the previously open question of whether every word-length function induces a compact quantum metric space; the answer is no."],"supporting_citations":[{"why":"Supplies the original construction of spectral triples from group length functions and the pseudometric on states that the paper tests.","marker":"[7]"},{"why":"Provides the definition of compact quantum metric space and Lip-norm that the theorem's conclusion negates.","marker":"[17]"},{"why":"Supplies the companion definition and framework for compact quantum metric spaces used in the paper.","marker":"[19]"},{"why":"Provides the total-boundedness criterion for a spectral triple to be a compact quantum metric space, which the proof invokes as Proposition 1.3.","marker":"[15]"},{"why":"Initiated the study of word-length spectral triples as compact quantum metric spaces and proved the positive integer lattice case that this paper's counterexamples do not extend.","marker":"[18]"},{"why":"Extends metricity to polynomial-growth groups, marking the positive territory outside of which the new counterexamples lie.","marker":"[6]"},{"why":"Explicitly records that no counterexample was known, the gap this paper fills.","marker":"[2]"}],"fun_headline_variants":["First counterexample: lamplighter triples not metric","Lamplighter word-length triples fail to be metric","No metric spectral triple for any lamplighter group","Lamplighter counterexample: spectral triples not metric"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof's load-bearing premise is the local estimate of Proposition 2.3, that toggling one lamp changes word length by at most a constant (depending on the generating set) times one plus the tree distance from the new lamp to the minimal subtree spanning the old configuration, the identity, and the base displacement; if that estimate failed for some finite symmetric generating set, the averaging elements would not be shown to lie in the unit Lipschitz ball and the separation argument would break down.","fun_headline_variants_meta":{"raw":{"variants":["First counterexample: lamplighter triples not metric","Lamplighter word-length triples fail to be metric","No metric spectral triple for any lamplighter group","Lamplighter counterexample: spectral triples not metric"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000805,"raw_usage":{"total_tokens":3559,"prompt_tokens":990,"completion_tokens":2569,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":606,"completion_tokens_details":{"reasoning_tokens":2503}},"tokens_in":606,"tokens_out":2569,"duration_ms":18413,"temperature":1.0,"reasoning_tokens":2503,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T00:18:56.726038+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take $d=2$ with the standard generators of the free group and compute the operator norm of $[D_\\ell,z_r]$ for growing $r$: Theorem A predicts a uniform upper bound, so any unbounded growth would disprove the main claim. Separately, one can hunt for a configuration $E,g,x\\notin E$ where the left-hand side of Proposition 2.3 exceeds any constant multiple of $1+d_T(x,H(E,g))$; such a configuration would directly falsify the key estimate on which the proof rests.","supporting_citations":[{"cited_title":"Connes,Compact metric spaces, Fredholm modules, and hyperfiniteness,Ergodic Theory Dynam","cited_arxiv_id":null,"evidence_quote":"Supplies the original construction of spectral triples from group length functions and the pseudometric on states that the paper tests."},{"cited_title":"Rieffel,Metrics on state spaces, Doc","cited_arxiv_id":null,"evidence_quote":"Provides the definition of compact quantum metric space and Lip-norm that the theorem's conclusion negates."},{"cited_title":"Rieffel,Gromov–Hausdorff distance for quantum metric spaces, Mem","cited_arxiv_id":null,"evidence_quote":"Supplies the companion definition and framework for compact quantum metric spaces used in the paper."},{"cited_title":"Ozawa, M","cited_arxiv_id":null,"evidence_quote":"Provides the total-boundedness criterion for a spectral triple to be a compact quantum metric space, which the proof invokes as Proposition 1.3."},{"cited_title":"Rieffel,Group C ∗-algebras as compact quantum metric spaces, Doc","cited_arxiv_id":null,"evidence_quote":"Initiated the study of word-length spectral triples as compact quantum metric spaces and proved the positive integer lattice case that this paper's counterexamples do not extend."},{"cited_title":"Christ, M","cited_arxiv_id":null,"evidence_quote":"Extends metricity to polynomial-growth groups, marking the positive territory outside of which the new counterexamples lie."}],"review_version":1}