{"id":"9ecd537d-9a5a-4d3e-9410-4b31b3ceb4a8","arxiv_id":"2412.12462","paper_version":2,"verdict":"REJECT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper proves equality of interleaving and bottleneck distances for pointwise finite-dimensional nilpotent representations of the continuous quiver of type tilde A.","lead":"This paper proves an Isometry Theorem for persistence modules on a circle (continuous quiver of type tilde A), stating that the interleaving distance equals the bottleneck distance between persistence diagrams. It extends a central stability result of persistent homology from real-line data to periodic, circle-valued data.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Proposition 4.3 asserts finiteness of persistence barcodes without proof; pointwise finite nilpotent tilde-A representations can have infinite barcodes, so Lemma 4.2 does not apply and Theorem 3.2 is unproved in stated generality.","rationale":"Both the reader and I identify the same critical gap: in Proposition 4.3 the authors need finite quotient barcodes to invoke Lemma 4.2, but finiteness is asserted, not proved, and is false for a natural class of pointwise finite-dimensional nilpotent representations. My concrete example shows that the theorem's hypotheses do not imply the technical condition on which the proof depends. I did not find a counterexample to the theorem itself; the equality may survive via approximation by finite barcodes. However, a proof that simply assumes the needed finiteness is not a proof of the stated theorem. The paper should either add a tameness/finiteness hypothesis to Theorem 3.2 or supply a limiting argument to remove it. Since the central claim is unsupported in its stated generality, the rejection stands. I agree with the reader that Lemma 4.2's proof also has gaps, but the finiteness issue alone is sufficient and is the most load-bearing.","tokens_in":9156,"tokens_out":30457,"duration_ms":282980,"concrete_test":"Construct V = ⊕_{n≥1} T_{|1/(n+1), 1/(n+1)+1/(n+1)^2|} and W = 0. Verify: (i) V is pointwise finite-dimensional and nilpotent; (ii) dgm(V) in R^2/∼ is infinite, so the hypotheses of Lemma 4.2 as invoked in Proposition 4.3 fail. Then compute both sides of Theorem 3.2: db = di = sup_n (length_n)/2 = 1/8. If equality holds, Theorem 3.2 may be true but needs an approximation argument; if not, it is false. Either outcome settles whether the missing finiteness condition is merely a proof gap or a genuine obstruction.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The proof of Theorem 3.2 reduces to Proposition 4.3, where after Theorem 2.1 the authors state: 'Since dgm(Ψ(V)) and dgm(Ψ(W)) are σ-invariant multisets such that dgm(Ψ(V)) and dgm(Ψ(W)) are finite, Lemma 4.2 implies...' Lemma 4.2 requires finite quotient multisets \\bar A and \\bar B. But pointwise finite-dimensional nilpotent representations of \\tilde A_R need not have finite barcodes. For example, V = ⊕_{n≥1} T_{|1/(n+1), 1/(n+1)+1/(n+1)^2|} is pointwise finite-dimensional (each S^1-fiber is covered by only finitely many summands) and nilpotent (every interval has length <1, so the monodromy around the circle is zero), yet its persistence diagram has one orbit-class per n, giving infinitely many classes. Hence the quoted finiteness assertion is false, or at best unproven, under the theorem's hypotheses. Without finiteness—or a limiting argument replacing Lemma 4.2—the inequality in Proposition 4.3, and therefore Theorem 3.2, does not follow. The theorem may still be true, but the proof as written does not cover the stated generality.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper aims to generalize the Isometry Theorem from continuous quivers of type A to continuous quivers of type \\tilde{A} (the circle). Using the Rock–Zhu equivalence between representations of \\tilde{A}_R and representations of a continuous quiver of type A with an automorphism (denoted Q), the authors claim that for pointwise finite-dimensional nilpotent representations V and W of \\tilde{A}_R, the bottleneck distance between their persistence diagrams in R^2/\\sim equals the interleaving distance on \\tilde{A}_R. The proof splits into two inequalities: Proposition 4.3 (stability, using the type-A theorem and a lemma comparing bottleneck distances on R^2 and R^2/\\sim) and Proposition 4.7 (converse, using a lemma on interleavings of interval representations).","tokens_in":9450,"tokens_out":7816,"duration_ms":68872,"significance":"If correct, the result would extend the isometry theorem to circle-valued persistence modules, a natural and potentially useful setting for topological data analysis and representation theory. The paper's strategy—importing the well-established type-A theorem and the Rock–Zhu equivalence—is attractive and would make the proof short. However, the proof as written contains significant gaps concerning infinite barcodes and a false assertion in Lemma 4.2. These issues are load-bearing: they affect both directions of the claimed isometry. The underlying idea is plausible, but the current manuscript does not establish Theorem 3.2 in its stated generality.","major_comments":[{"comment":"The assertion that dgm(Ψ(V)) and dgm(Ψ(W)) are finite is unjustified and in fact false in general under the theorem's hypotheses. Pointwise finite-dimensional nilpotent representations of \\tilde{A}_R can have infinite barcodes. For example, take V = ⊕_{n≥1} T_{[1/(n+1), 1/(n+1)+1/(n+1)^2]}. This is pointwise finite-dimensional because for any circle coordinate, only finitely many intervals cover it, and it is nilpotent because every interval has length less than 1. Its persistence diagram has one orbit class per n, so the quotient multiset in R^2/\\sim is infinite. Thus Lemma 4.2, which requires finite quotient multisets, cannot be applied. Without a limiting argument that replaces Lemma 4.2 or an additional finiteness hypothesis in Theorem 3.2, the inequality in Proposition 4.3—and hence Theorem 3.2—is not proved in the stated generality.","section":"§4.2, Proposition 4.3"},{"comment":"The proof of Lemma 4.2 contains a false step: 'Since c(P) is finite, we have |\\bar{B}''| ≥ |\\bar{A}'|.' Because A and B are σ-invariant, each orbit is infinite, so a single orbit of B can absorb matches from many different orbits of A. Concretely, let A have two orbits \\bar{a}_1, \\bar{a}_2 and B have one orbit \\bar{b}. Match the n-th point of \\bar{a}_1 to a point in \\bar{b} with offset ε, and the n-th point of \\bar{a}_2 to another point in \\bar{b} with offset 1/2+ε, choosing distinct points in \\bar{b}. Then all points in both \\bar{a}_1 and \\bar{a}_2 are matched with bounded cost, so |\\bar{A}'|=2 while |\\bar{B}''|=1, contradicting the claim. Consequently inequality (4.1) is not established. The lemma requires a correct proof, or the paper must find another way to relate bottleneck distances on R^2 and R^2/\\sim for σ-invariant multisets.","section":"§4.2, Lemma 4.2"},{"comment":"The proof of Lemma 4.6 is sketched as 'Similarly to the proof of the converse stability part of Theorem 2.1 in [17]' and then uses a direct-sum decomposition with a partial matching of summands. For infinite barcodes this is not justified: one needs to show that an interleaving of infinite direct sums can be assembled from interleavings of matched summands, that the maximum (or supremum) of distances over an infinite matching is attained or handled by limits, and that unmatched intervals at the boundary are controlled. This is the same infinity issue raised in Proposition 4.3; as written, the converse inequality di_{\\tilde{A}_R}(V,W) ≤ db_{R^2/\\sim}(dgm(V),dgm(W)) is also not fully proved for the stated class of representations.","section":"§4, Lemma 4.6"}],"minor_comments":[{"comment":"In the statement of Theorem 3.2, 'db,, R2/ ∼' contains a double comma; it should read 'db_{R^2/\\sim}'.","section":"§3.2, Theorem 3.2"},{"comment":"In Equation (4.8), the expression 'db_{R^2}(dgm(Ψ(V)), Ψ(dgm(W)))' appears to have a typo; it should be 'db_{R^2}(dgm(Ψ(V)), dgm(Ψ(W)))'.","section":"§4.2, Proposition 4.3, Eq. (4.8)"},{"comment":"The definition of the bottleneck cost on R^2/\\sim uses the term sup_{\\bar{s}=(s_x,s_y)∈S} (s_x−s_y)/2, which is not well-defined because the value depends on the chosen representative of the equivalence class \\bar{s}. The authors should specify that the infimum over representatives (or an invariant representative, e.g., with first coordinate in [0,1)) is taken.","section":"§3.2, bottleneck distance definition"},{"comment":"The interval representation T_{|a,b|} of Q introduced as ⊕_{k∈Z}(σ^*)^k(T_{|a,b|}) uses the same notation as the interval representation of \\tilde{A}_R from §3.1. This notational collision is confusing, especially because Ψ(T_{|a,b|}) = T_{|a,b|} is claimed; consider using distinct notations for the two types of interval representations.","section":"§4.1"}],"recommendation":"major_revision","confidential_remarks":"The paper is a short note that reduces the circle-valued isometry theorem to the known type-A case. The reduction is natural and the statement is plausible, but the current proof has two serious gaps: the unproved finiteness of persistence diagrams under the stated hypotheses, and an invalid cardinality step in Lemma 4.2. The authors should be given the opportunity to repair these, either by adding an explicit finiteness assumption or by proving a limiting version of Lemma 4.2 and handling infinite barcodes in Lemma 4.6. If these repairs succeed, the paper would be a useful contribution; as it stands, the central claim is not established."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a reasonable short paper that wants to extend the Isometry Theorem to circle-valued persistence (type tilde A). The main theorem is new in the literature and, if true, would be useful for TDA on circle-parameterized data. The proof route -- go through Rock-Zhu equivalence to type A with automorphism, then reduce to the classical Isometry Theorem -- is sensible and mostly clearly explained.\n\nWhat's good: the paper states the theorem cleanly, cites the right prior work (Oudot, Hanson-Rock, Rock-Zhu), and the lemmas for indecomposable intervals (4.5, 4.6) look fine. The reduction in Proposition 4.3 is structurally sound.\n\nThe problem is in the middle of the proof. Proposition 4.3 asserts without proof that dgm(Ψ(V)) and dgm(Ψ(W)) are finite. That is not a consequence of the hypotheses. A pointwise finite nilpotent representation of tilde A can absolutely have infinitely many interval summands with distinct orbit classes -- e.g. take intervals of length 1/(n+1)^2 for each n in a disjoint family near 0. This gives an infinite barcode, and there is no reason to think Ψ sends it to a finite multiset. Lemma 4.2 explicitly requires finite quotients, so Proposition 4.3 cannot invoke it.\n\nThere is also a separate issue inside Lemma 4.2's proof: the line \"Since c(P) is finite, we have |B''| >= |A'|\" does not follow. Finite cost controls the distance of matched pairs and the spread of unmatched points; it doesn't bound the number of orbit classes. A sigma-invariant matching can match points from two different A-orbits into the same B-orbit, so the cardinality inequality can fail.\n\nThese aren't cosmetic. They are the load-bearing steps that connect the tilde-A distance to the quotient bottleneck. The theorem itself might be true, and a repair may be straightforward -- for example, add a finiteness or tameness hypothesis on the barcodes, or prove a limiting version of Lemma 4.2 for infinite multisets. But the proof as written does not establish Theorem 3.2.\n\nWho is this for: researchers in persistent homology who care about circle-valued data. A serious referee should see it. I'd suggest sending to review with the expectation of major revision rather than desk rejecting, because the statement is natural and the gap is identifiable rather than hopeless.","headline":"The paper's main theorem is plausible and the topic is timely, but the proof has a load-bearing finiteness gap in Proposition 4.3, and Lemma 4.2 contains a false cardinality claim.","tokens_in":9940,"tokens_out":3830,"would_cite":false,"duration_ms":32955,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["16G20","55N31"],"pacs":[],"model":"deepseek-v4-flash","headline":"The Isometry Theorem for persistence extends from the real line to the circle.","keywords":["continuous quiver","isometry theorem","persistence diagram","bottleneck distance","interleaving distance","nilpotent representation","circle-valued persistence","quiver representation"],"falsifier":"Take $V=\\bigoplus_{n\\in\\mathbb{N}} T_{|n,\\ n+1/(n+2)|}$ and $W=0$: $V$ is pointwise finite-dimensional and nilpotent, but its barcode has infinitely many distinct classes in $\\mathbb{R}^2/\\sim$, so Proposition 4.3's finiteness assertion is false and Lemma 4.2 does not apply. Computing $d_{b,\\mathbb{R}^2/\\sim}(\\operatorname{dgm}(V),\\operatorname{dgm}(W))$ and $d_{i,\\tilde A_R}(V,W)$ for this pair (or a pair of two such sums) settles whether the stated theorem still holds for infinite quotient diagrams or whether the proof needs an additional hypothesis.","tokens_in":8942,"feed_emoji":"⭕","tokens_out":15453,"duration_ms":136765,"temperature":0.7,"pith_summary":"The paper's goal is to prove an Isometry Theorem for the continuous quiver $\\tilde A_R$, whose vertices are points of $\\mathbb{R}$ modulo integer shifts, i.e. a circle. The main theorem says that for pointwise finite-dimensional nilpotent representations $V$ and $W$ of this quiver, the bottleneck distance between their persistence diagrams in $\\mathbb{R}^2/\\sim$ equals the interleaving distance between the representations. If correct, this carries the central stability result of persistent homology from the real line to circle-valued persistence modules, so periodic and angular data are covered by the same distance identity. The proof works by transporting representations through a category equivalence to a continuous quiver of type $A$ with an automorphism, where the classical Isometry Theorem and a comparison lemma for quotient bottleneck distances complete the argument.","feed_headline":"Circle persistence modules obey the isometry theorem","feed_subtitle":"Bottleneck distance between barcodes equals interleaving distance for nilpotent modules on the circle.","key_machinery":"The load-bearing device is the category equivalence $\\Psi$ between representations of $\\tilde A_R$ and representations of a continuous quiver of type $A$ with automorphism $Q=(A_{\\mathbb{R}},\\sigma)$, where $\\sigma(x)=x+1$. Interval representations of $\\tilde A_R$ become direct sums over $\\mathbb{Z}$ of translates of ordinary interval representations, and $\\Psi$ preserves interleaving distances. The proof then combines the classical Isometry Theorem for $A_{\\mathbb{R}}$ with Lemma 4.2, which compares the bottleneck distance of a $\\sigma$-invariant multiset in $\\mathbb{R}^2$ with the bottleneck distance of its quotient in $\\mathbb{R}^2/\\sim$, assuming the quotient multiset is finite. Lemmas 4.4 through 4.6 transfer the standard interleaving estimates for single intervals to the quotient setting, yielding both inequalities.","core_discovery":"The central claim (Theorem 3.2) is exact equality $$d_{b,\\mathbb{R}^2/\\sim}(\\operatorname{dgm}(V),\\operatorname{dgm}(W)) = d_{i,\\tilde A_R}(V,W)$$ for every pair of pointwise finite-dimensional nilpotent representations $V,W$ of $\\tilde A_R$. The persistence diagram $\\operatorname{dgm}(V)$ is the multiset in $\\mathbb{R}^2/\\sim$ formed by the equivalence classes of the endpoint pairs $(a_i,b_i)$ in the interval decomposition $V=\\bigoplus_i T_{|a_i,b_i|}$. The intended reading is that two circular persistence modules are $\\varepsilon$-interleaved exactly when their barcodes admit a partial matching whose bottleneck cost is at most $\\varepsilon$. This is the direct analogue, for the circle, of the classical Isometry Theorem for the continuous quiver of type $A$.","pith_inferences":["The proof does not establish finiteness of the quotient persistence diagram; because pointwise finite-dimensionality alone need not force finitely many $\\mathbb{Z}$-orbits of barcode endpoints, the theorem as stated likely requires an additional approximation argument for infinite quotient diagrams.","The same quotient construction should work for other discrete translation groups acting on $\\mathbb{R}$, giving isometry theorems for persistence modules on other periodic quotient spaces, provided a finite-barcode condition holds.","A direct computation of both distances for a pair with accumulating barcode endpoints would show whether the finiteness assumption is merely a proof artifact or a genuine restriction on the validity of the theorem."],"forward_implications":["The interleaving distance between two pointwise finite-dimensional nilpotent representations of $\\tilde A_R$ is completely determined by their persistence diagrams.","Stability holds in both directions: a bottleneck distance at most $\\varepsilon$ between diagrams is equivalent to an $\\varepsilon$-interleaving, so small changes in barcodes cannot hide large structural changes in the modules.","Circle-valued persistence modules, which arise in data with angular or periodic coordinates, now satisfy the same distance identity that was previously known only for modules indexed by the real line.","Bottleneck distance on $\\mathbb{R}^2/\\sim$ provides an intrinsic metric on isomorphism classes of these representations, since it coincides with the interleaving distance defined directly from morphisms."],"supporting_citations":[{"why":"It establishes the category equivalence between representations of $\\tilde A_R$ and continuous quivers of type $A$ with automorphism, which transfers the distance computation.","marker":"[18]"},{"why":"It supplies the interval decomposition theorem for pointwise finite-dimensional nilpotent representations of $\\tilde A_R$, on which the persistence diagram is defined.","marker":"[13]"},{"why":"It contains the Isometry Theorem for type $A$ and the interval interleaving estimates that Lemma 4.6 adapts to the automorphism setting.","marker":"[17]"},{"why":"It introduces continuous quivers of type $A$ and their interval representations, which are the building blocks used to construct the $\\tilde A_R$ representations.","marker":"[14]"}],"fun_headline_variants":["Bottleneck equals interleaving for circle persistence modules","Circle persistence: bottleneck distance equals interleaving distance","Isometry for circle: bottleneck = interleaving","Bottleneck distance = interleaving distance on the circle"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof assumes, without proof, that every pointwise finite-dimensional nilpotent representation has a persistence diagram with only finitely many classes after identifying barcode endpoints that differ by an integer shift; this finiteness is needed to apply the quotient bottleneck comparison lemma, and the argument gives no reason to think it always holds.","fun_headline_variants_meta":{"raw":{"variants":["Bottleneck equals interleaving for circle persistence modules","Circle persistence: bottleneck distance equals interleaving distance","Isometry for circle: bottleneck = interleaving","Bottleneck distance = interleaving distance on the circle"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001013,"raw_usage":{"total_tokens":4176,"prompt_tokens":738,"completion_tokens":3438,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":354,"completion_tokens_details":{"reasoning_tokens":3372}},"tokens_in":354,"tokens_out":3438,"duration_ms":20615,"temperature":1.0,"reasoning_tokens":3372,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T14:05:36.749110+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take $V=\\bigoplus_{n\\in\\mathbb{N}} T_{|n,\\ n+1/(n+2)|}$ and $W=0$: $V$ is pointwise finite-dimensional and nilpotent, but its barcode has infinitely many distinct classes in $\\mathbb{R}^2/\\sim$, so Proposition 4.3's finiteness assertion is false and Lemma 4.2 does not apply. Computing $d_{b,\\mathbb{R}^2/\\sim}(\\operatorname{dgm}(V),\\operatorname{dgm}(W))$ and $d_{i,\\tilde A_R}(V,W)$ for this pair (or a pair of two such sums) settles whether the stated theorem still holds for infinite quotient diagrams or whether the proof needs an additional hypothesis.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It establishes the category equivalence between representations of $\\tilde A_R$ and continuous quivers of type $A$ with automorphism, which transfers the distance computation."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It supplies the interval decomposition theorem for pointwise finite-dimensional nilpotent representations of $\\tilde A_R$, on which the persistence diagram is defined."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It contains the Isometry Theorem for type $A$ and the interval interleaving estimates that Lemma 4.6 adapts to the automorphism setting."},{"cited_title":"Igusa, J","cited_arxiv_id":null,"evidence_quote":"It introduces continuous quivers of type $A$ and their interval representations, which are the building blocks used to construct the $\\tilde A_R$ representations."}],"review_version":1}