{"id":"51e238ed-5538-4d65-8beb-9b58b7341130","arxiv_id":"2608.13526","paper_version":1,"verdict":"CONDITIONAL","confidence":"LOW","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"On any graph manifold, all transitive pseudo-Anosov flows have at most finitely many orbit-equivalence classes after lifting to a single finite cover, and a new complete bar code invariant decides when glued flows are orbit equivalent.","lead":"This paper proves a virtual finiteness theorem: on any graph manifold, transitive pseudo-Anosov flows fall into finitely many orbit equivalence classes after lifting to a single finite cover. It also introduces a new combinatorial invariant, the complete bar code, that characterizes when two flows built by gluing the same pieces are orbit equivalent.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Virtual finiteness rests on Lemma 7.5, whose key skewness assertion is deferred to the unpublished [BLMP]; the sketched axis argument does not establish the claim.","rationale":"The reader's weakest-assumption analysis identifies Lemma 7.5 as the load-bearing dependency, and I agree. The entire Section 7 chain—Lemma 7.4, Lemma 7.5, Proposition 7.7, and the final contradiction in Section 7.3—uses Lemma 7.5 to conclude that the relevant pieces and doubled flows are skew, so that the published finiteness result [Mar25, Corollary C] can be applied. The paper itself flags the [BLMP] dependency in the proof of Lemma 7.5, and the sketched argument does not supply the missing proof: the existence of the axes A_j in the stable leaf space is precisely what [BLMP] is supposed to establish, and without it the conclusion that the doubled flow is R-covered and skew is unsupported. This does not mean the theorem is false; much of the paper, especially the orbit-space machinery in Sections 3–6, appears carefully built and may well be correct. But the current manuscript does not contain a complete proof of the key finiteness mechanism, so the conditional verdict is appropriate and no verdict change is needed.","tokens_in":48392,"tokens_out":16095,"duration_ms":172007,"concrete_test":"Replace the [BLMP] citation in Lemma 7.5 with a complete proof. Concretely, from the hypotheses of Convention 7.3, verify: (i) the hypotheses of [Pau25, Lemma 11.4] and [Pau25, Theorem 1] for the double of (P, φ); (ii) that each vertex group π1(Q_j) of the doubled manifold preserves a unique axis A_j in the stable leaf space of the resulting Anosov flow, and that for adjacent pieces sharing a boundary torus T, the subgroup π1(T) contains an element acting freely on the orbit space so that A_i = A_j; (iii) that Λ(Q^s) = R and φ' is not a suspension flow. If any of these steps requires a statement not present in [Pau25], [Mar25], or the rest of this paper, the dependency on [BLMP] is unresolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The proof of Lemma 7.5 in Section 7.1 states that the Anosov flow φ' on the doubled manifold M' is skew, and the proof is 'proved in [BLMP]', an unpublished manuscript. The sketch that follows assumes that each vertex group π1(Q_j) preserves an axis A_j in the stable leaf space of φ'; this is exactly the content of [BLMP], not a consequence of the published [Pau25] or [Mar25]. If Lemma 7.5 fails, then in Proposition 7.7 the pieces (P_i, φ_i) cannot be shown to be skew, and the appeal to [Mar25, Corollary C] for finitely many skew Anosov flows on a fixed manifold has no basis. The contradiction argument in Section 7.3 then collapses, so Theorems 7.2, F, G, and B are conditional on an unavailable proof. This is a genuine external dependency, not an internal inconsistency, but it is load-bearing: without Lemma 7.5, the paper supplies no finiteness mechanism for the non-periodic pieces.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies pseudo-Anosov flows on 3-manifolds obtained by gluing flow pieces along tori. It introduces a finite combinatorial invariant, the complete bar code, encoding the traces of the stable and unstable foliations on the cutting tori. The main theorems assert that equivalent gluing triples with the same complete bar code induce flows with equivalent free homotopy data, and orbit-equivalent flows when transitivity is assumed (Theorems C and A); that only finitely many complete bar codes occur in an equivalence class (Proposition D); and a virtual finiteness statement for transitive pseudo-Anosov flows on graph manifolds (Theorem B), deduced from the more general Theorem 7.2 / Theorem F. The arguments use the orbit-space machinery of Barbot and Fenley and earlier work of the same authors, with a final step that depends on a lemma whose proof is deferred to an unpublished reference.","tokens_in":48606,"tokens_out":10433,"duration_ms":98606,"significance":"If the deferred lemma is supplied, the paper would constitute a substantial advance: a general gluing-uniqueness criterion and a virtual resolution of a case of the Finiteness Conjecture for graph manifolds, a problem on Kirby's list. The complete bar code is a concrete and well-motivated finite invariant, and Theorem E provides a reusable criterion for equality of free homotopy data. The paper is carefully organized and credits prior work appropriately. Its main weakness is that the finiteness theorems rest on an unpublished lemma (Lemma 7.5), so the central claim is not yet self-contained; there are also two further technical points that need repair or fuller proof.","major_comments":[{"comment":"Lemma 7.5 is load-bearing for the virtual finiteness theorem: it asserts that a pseudo-Anosov piece obtained by gluing skew pieces is itself skew. The proof states that 'the fact that the Anosov flow φ′ on M′ is skew is proved in [BLMP]' and then gives a sketch whose starting point — that each vertex group π1(Q_j) preserves an axis A_j in the stable leaf space of φ′ — is exactly the content of [BLMP], not a consequence of the published [Pau25] or [Mar25]. Proposition 7.7 needs Lemma 7.5 to classify the non-periodic pieces as skew and then apply [Mar25, Corollary C], so Theorems 7.2, F, G, and B are conditional on unpublished work. The authors should either include a complete proof of Lemma 7.5 in this paper or restate the finiteness theorems as conditional on the appearance of [BLMP]; as it stands, the finiteness claim is not self-contained.","section":"§7.1, Lemma 7.5"},{"comment":"Lemma 2.19 is used to prove Proposition 2.17, which supplies the finite cover with embedded modified JSJ tori in Theorem 7.2. The proof has two gaps. In Step 3, the word β obtained from a subarc of c is asserted to lie in the forbidden set F, but membership in F requires β to be a reduced word in distinct letters β_i; the text does not justify that a subarc of a simple closed curve whose retraction is non-simple cannot repeat a letter. In Step 4, the claim that taking successive finite covers for different fatgraphs does not introduce new self-intersections is asserted without proof. Since Proposition 2.17 is needed for the finite-cover passage, this lemma needs a complete proof or a precise reference.","section":"§2.5, Lemma 2.19"},{"comment":"Theorem A is advertised in the introduction as the characterization theorem, but neither Section 6 nor Section 7 contains an explicit proof of it. Theorem C supplies the 'if' direction for equivalent triples with the same complete bar code (and hence, after cutting along the tori, for the flows in Theorem A). The converse — that an orbit equivalence between φ1 and φ2 forces the corresponding triples to have the same complete bar code — is not proved or cited. This is the uniqueness half of the paper's main statement and should be proved explicitly, or the statement should be amended to a one-way implication if the converse is intended to be immediate from the definitions.","section":"§6, Theorem A"}],"minor_comments":[{"comment":"The phrase 'same complete bar code' for two flows in Theorem A is not formally defined; it is defined only for pseudo-Anosov triples. Please spell out how the definition is transferred to flows via the modified JSJ decomposition.","section":"Definitions 4.20 and 4.28"},{"comment":"The proof of Proposition D leaves several claims unproved, in particular that the projection of a stable cylinder to the base surface is an embedded arc crossing at most one vertex of the fatgraph, and that the isotopy class of these projected arcs is fixed. Since Corollary 6.11 depends on Proposition D, this part should be expanded.","section":"§6, Proposition D"},{"comment":"There is a typo in the proof: 'an-Dehn twist' should be 'a Dehn twist'.","section":"§2.2, Proposition 2.7"},{"comment":"The reference [BLMP] is listed as 'in preparation'. If Lemma 7.5 remains deferred, the paper should state explicitly which results are conditional on this unpublished work, preferably in the introduction.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper is a serious contribution and well within the journal's scope. The main concern is the unpublished reference [BLMP], which carries a load-bearing lemma for the finiteness theorems; if the authors cannot supply the proof, the editors will need to decide whether a conditional result is acceptable. The paper also leans heavily on earlier work of the same authors, but most of that work is published; the only problematic citation is [BLMP]. I recommend major revision with a request for a complete proof of Lemma 7.5 and a repair of Lemma 2.19."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Martin,\n\nHere's my read on the Barthelmé–Paulet preprint. The core content is genuinely new: the complete bar code gives a finite combinatorial invariant that characterizes orbit equivalence of transitive pseudo-Anosov flows obtained by gluing pieces along quasi-transverse tori (Theorems A/C). Theorem E, the order-preservation criterion for equality of free homotopy data, is a clean and useful generalization of earlier work by the same authors. Proposition D, finiteness of complete bar codes per equivalence class, is also new. These results are carefully argued and I see no circularity: the invariants are defined directly from the flows and the proofs don't assume their own conclusion.\n\nThe soft spot is exactly where the stress-test note lands. Lemma 7.5, used to prove Theorem 7.2 and hence Theorems B, F, and G, says that a piece obtained by gluing skew pieces is itself skew. The proof in Section 7.1 defers the key fact to the unpublished manuscript [BLMP]. The sketched axis argument assumes each vertex group π1(Q_j) preserves an axis in the stable leaf space; that is precisely what [BLMP] is supposed to establish. So the virtual finiteness theorem is conditional on an unavailable proof. This is not an internal contradiction, but it is load-bearing: without Lemma 7.5, the finiteness mechanism for non-periodic pieces collapses. It is a real dependency, and the authors should either supply a complete proof or state the result as conditional.\n\nOther technical parts, for example Lemma 2.19 on finite covers with embedded modified JSJ tori, look plausible but I haven't checked every step. The paper is honest about these dependencies, and the citation pattern is appropriate: heavy reliance on earlier work by the same group, but those papers are published and independent.\n\nWho should read it? Researchers working on pseudo-Anosov flows and 3-manifold topology. The classification results (A/C/E) stand on their own, and the finiteness theorem is a major step if the gap is filled. I would bring it to a reading group, and I would cite it for the bar code results even before the gap is closed.\n\nRecommendation for review: send it out—this is a serious paper that deserves referee time. But the referee should demand a complete proof of Lemma 7.5 or a clear statement that the finiteness theorems depend on an unpublished companion. With that, accept; without it, major revision.\n\nBest","headline":"Strong paper with a real gap: the main finiteness theorem depends on a lemma deferred to an unpublished companion paper.","tokens_in":49134,"tokens_out":3273,"would_cite":true,"duration_ms":32276,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["37D20","57K30"],"pacs":[],"model":"deepseek-v4-flash","headline":"Two pseudo-Anosov flows on the same toroidal 3-manifold are orbit equivalent exactly when their complete bar codes agree, and this yields virtual finiteness of transitive pseudo-Anosov flows on graph manifolds.","keywords":["pseudo-Anosov flows","graph manifolds","orbit equivalence","gluing","bar code","free homotopy data","JSJ decomposition","virtual finiteness"],"falsifier":"Build a graph manifold $M$ and a family of transitive pseudo-Anosov flows such that, for every finite cover of $M$, the lifted flows represent infinitely many orbit-equivalence classes; Theorem B would be false. A cheaper target is to exhibit a non-skew Anosov flow obtained by gluing two skew pieces along quasi-transverse tori, which would falsify Lemma 7.5 and with it the mixed-manifold theorem.","tokens_in":1742,"feed_emoji":"🧩","tokens_out":1773,"duration_ms":79582,"temperature":0.7,"pith_summary":"The paper answers the long-open gluing question for pseudo-Anosov flows: if two flows are built from the same pieces along tori, they are orbit equivalent precisely when a finite combinatorial invariant, the complete bar code, matches. It then uses this uniqueness result to prove a virtual version of the Finiteness Conjecture: on any graph manifold, any family of transitive pseudo-Anosov flows becomes finite up to orbit equivalence after lifting to one finite cover whose degree depends only on the topology of the manifold. This matters because the full Finiteness Conjecture remains open, and the result reduces a dynamical classification problem to finite combinatorial data attached to the JSJ decomposition.","feed_headline":"Same bar code, same pseudo-Anosov flow","feed_subtitle":"Matching bar codes force orbit equivalence; graph-manifold finiteness follows.","key_machinery":"The complete bar code of a pseudo-Anosov triple $(P,\\varphi,f)$ is a finite combinatorial datum: on each parallel boundary torus it records the cyclic sequence of compact leaves of the boundary foliation, labeled stable, unstable, marked, or tangent, together with a shift locating the image of the gluing foliation. It extends the bar code of the boundary prefoliation, which is intrinsic to the piece, by including extra compact leaves created by the gluing; this is exactly the data needed to compare the traces of the stable and unstable foliations on the gluing tori. In the orbit space, a quasi-transverse torus becomes a chain of lozenges, and the partial order $\\prec$ between maximal lines of lozenges detects which orbit segments can travel from one boundary lift to another; preserving $\\prec$ is the condition that yields equality of free homotopy data and, for transitive flows, orbit equivalence.","core_discovery":"The central discovery is that gluing ambiguity is fully controlled by the boundary combinatorics of the stable and unstable foliations. For two transitive pseudo-Anosov flows on an orientable toroidal 3-manifold that already agree piecewise on a JSJ-like decomposition, the flows are orbit equivalent if and only if their complete bar codes agree. The complete bar code records, on each gluing torus, the cyclic order of compact leaves of the stable and unstable traces together with the periodic orbits where the foliations are tangent, up to a choice of first leaf. In the universal cover this data becomes a partial order on maximal lines of lozenges in the orbit space; preserving that order preserves free homotopy data, and for transitive flows free homotopy data determine orbit equivalence. Iterating the finiteness of bar codes in each equivalence class of a pseudo-Anosov triple over the finitely many JSJ pieces yields the virtual finiteness theorem.","pith_inferences":["The complete bar code together with the order $\\prec$ could yield an algorithmic recognition procedure for orbit equivalence, since both objects are finite and the order has a dynamical reformulation in terms of existence of orbit segments.","If the unpublished skew-gluing lemma used for mixed manifolds is established independently, the remaining finiteness argument is topological, suggesting virtual finiteness may hold for all toroidal 3-manifolds without the skewness hypothesis.","The proof structure indicates that the same finiteness should hold for non-transitive families if the conclusion is phrased in terms of free homotopy data rather than orbit equivalence; the transitive hypothesis is used only to convert equal free homotopy data into orbit equivalence.","A positive answer to the paper's Question 3 would upgrade the finite-cover statement to genuine finiteness of pseudo-Anosov flows on graph manifolds."],"forward_implications":["If two transitive pseudo-Anosov flows on a toroidal manifold agree on each JSJ piece, then checking a finite combinatorial invariant, the complete bar code, decides global orbit equivalence.","On a fixed graph manifold, every family of transitive pseudo-Anosov flows has at most $n$ orbit-equivalence classes after lifting to one finite cover whose degree depends only on the manifold.","For a fixed manifold, there are finitely many free homotopy data among flows built from equivalent pseudo-Anosov triples whose gluing returns that manifold.","A Dehn twist along a quasi-transverse torus in the direction of a compact leaf is a self-orbit equivalence of a transitive pseudo-Anosov flow.","The finiteness result extends to manifolds with mixed JSJ decomposition provided the flow is skew on each hyperbolic piece, and under orientability plus embedded-torus hypotheses the finite cover is unnecessary."],"supporting_citations":[{"why":"Supplies the classification and rigidity of totally periodic pseudo-Anosov flows in graph manifolds, used for periodic Seifert pieces.","marker":"[BF15]"},{"why":"Provides the description of pseudo-Anosov flows in free Seifert pieces, used to control compact leaves on boundary tori.","marker":"[BF21]"},{"why":"Gives the earlier orbit-equivalence and free-homotopy-data criteria that Theorem E generalizes.","marker":"[BFM25a]"},{"why":"Supplies the result that transitive pseudo-Anosov flows with equivalent free homotopy data are orbit equivalent, aside from a scalloped sign case.","marker":"[BFM25b]"},{"why":"Provides the building-block formalism for gluing flows with quasi-transverse boundary, used throughout the finiteness proof.","marker":"[Pau25]"},{"why":"Yields the finiteness of skew Anosov flows on a fixed manifold, a key input for the skew-piece step.","marker":"[Mar25]"},{"why":"Gives the classification of homeomorphisms that are Dehn twists on the boundary, used to make boundary orbit equivalences isotopic to the identity.","marker":"[McC06]"},{"why":"Supplies the unpublished result that gluing skew pieces produces a skew flow, on which Lemma 7.5 and the mixed-manifold theorem depend.","marker":"[BLMP]"}],"fun_headline_variants":["Bar codes determine orbit equivalence of glued flows","Virtual finiteness proven for pseudo-Anosov flows on graph manifolds","Gluing flows uniquely: bar codes are the key","Bar codes force orbit equivalence, yielding virtual finiteness"],"cache_read_input_tokens":51328,"weakest_assumption_plain":"The proof of the mixed-manifold extension relies on an unpublished claim that gluing skew pieces always gives a skew flow, and if that claim fails the finiteness theorems no longer follow.","fun_headline_variants_meta":{"raw":{"variants":["Bar codes determine orbit equivalence of glued flows","Virtual finiteness proven for pseudo-Anosov flows on graph manifolds","Gluing flows uniquely: bar codes are the key","Bar codes force orbit equivalence, yielding virtual finiteness"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000446,"raw_usage":{"total_tokens":2178,"prompt_tokens":792,"completion_tokens":1386,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":408,"completion_tokens_details":{"reasoning_tokens":1320}},"tokens_in":408,"tokens_out":1386,"duration_ms":11536,"temperature":1.0,"reasoning_tokens":1320,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T05:02:24.419798+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Build a graph manifold $M$ and a family of transitive pseudo-Anosov flows such that, for every finite cover of $M$, the lifted flows represent infinitely many orbit-equivalence classes; Theorem B would be false. A cheaper target is to exhibit a non-skew Anosov flow obtained by gluing two skew pieces along quasi-transverse tori, which would falsify Lemma 7.5 and with it the mixed-manifold theorem.","supporting_citations":[],"review_version":1}