{"id":"63c32db7-15bb-4f02-bf40-0dbb159c3c21","arxiv_id":"2505.21113","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"For every even n≥4, infinitely many surgeries on the n-chain link are hyperbolic L-spaces with n orbit-inequivalent pseudo-Anosov flows and n universally tight non-contactomorphic contact structures.","lead":"This paper constructs 3-manifolds that each carry many completely different pseudo-Anosov flows, and uses them to answer open questions about flows and contact structures. The examples are hyperbolic L-spaces with no taut foliations, so they show that even 'simple' spaces can host rich dynamics.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The construction hinges on c_{B0}(h) = -1/4 (Prop. 3.1); if this FDTC is wrong, the prong-counting argument for orbit inequivalence fails.","rationale":"The reader and I converge on the same load-bearing step: the fractional Dehn twist coefficient c_{B0}(h) = -1/4 in Proposition 3.1. It is the only place where the construction produces a surgery core whose prong count grows like M^{k+1}, and Proposition 3.4 needs these counts to be distinct and maximal to conclude orbit inequivalence. If c_{B0} were 0, then d_0 would be a meridian multiple and the distinguished core would have at most 4(n+1) prongs, so all prong counts would be < M and the rotation flows would not be distinguished. The proof as written reduces the computation to the quotient by tau and to a diagrammatic assertion about arcs alpha and beta; Figures 3-5 are not visible in the text, so the verification is not fully transparent. I found no other internal inconsistency: the surgery triangle proof in Appendix A is sound, the homology and hyperbolicity bounds are standard, and the contact-structure applications are clearly delegated to published results plus a detailed extension in Proposition 2.3. The FDTC claim is finite and effectively computable, and Remark 3.6 indicates an invariant train track computation exists. This is exactly a condition for full acceptance rather than a reason to reject.","tokens_in":18019,"tokens_out":12354,"duration_ms":148639,"concrete_test":"Compute c_{B0}(h) directly for the explicit monodromy h = D^{-1}_{b0} o D_{b1} o ... o D^{-1}_{b_{n-1}} o D^{-1}_a by implementing the mapping class in a program such as flipper, or by running the HKM right-veering/FDTC algorithm on the invariant train track mentioned in Remark 3.6, for at least n=4 and n=6. If the returned value at B0 is -1/4 and all other c_{Bi} are 0, Proposition 3.1 is confirmed; if it differs, the prong-counting argument in Proposition 3.4 collapses and Theorem 1.1 is unproved as written.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The proof of orbit inequivalence in Proposition 3.4 distinguishes the flows by the unique maximal prong count of the surgery core gamma_0, namely l_0(M^{k+1}+24), which exceeds every other prong count (bounded by M). This equality comes from Lemma 3.3, whose degeneracy slope d_0 = l_0(-6mu_0+lambda_0) depends on c_{B0}(h) = -1/4 in Proposition 3.1. If this FDTC were instead 0, then d_0 would be a multiple of mu_0 and the surgical distance at component 0 would be at most 4(n+1) < M, so no orbit would have a uniquely maximal prong count and the rotation flows would not be distinguished by this argument. The computation of c_{B0}(h) is carried out by passing to the quotient by tau and asserting that c_{\\bar B0}(\\bar h^2 o D_delta) = 0 because the arcs alpha and beta in Figure 5 are sent respectively right and left at \\bar B0. This is a diagrammatic check, not an explicit algebraic computation; Figures 3-5 are not reproduced in the text, and Remark 3.6 mentions an invariant train track computation that would settle it but does not present it. Thus a single, localized, checkable FDTC computation is the load-bearing step for the main theorem.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves that for each even n ≥ 4 there are infinitely many quarter-integer Dehn surgeries on the n-component chain link L_n that are hyperbolic L-spaces and carry n pairwise orbit-inequivalent pseudo-Anosov flows. The flows are constructed by Fried surgery from a fibered link L'_n = -L_0 ∪ L_1 ∪ ... ∪ L_{n-1}, whose monodromy is an explicit product of Dehn twists. The key invariant distinguishing the flows is the maximal prong count of the surgery core γ_0, which is computed from the fractional Dehn twist coefficient c_{B_0}(h) = -1/4 and the bounds from Lemma 3.2. The authors then use standard L-space and hyperbolicity results to show the surgered manifolds have odd-order first homology, are hyperbolic and are L-spaces, hence have no taut foliations. They also show the flows have no perfect fits, hence are quasigeodesic, and, after passing to orientation-reversed manifolds, that they give rise to universally tight contact structures; using cylindrical contact homology and the Barthelmé–Frankel–Mann reconstruction theorem, they conclude the contact structures are pairwise non-contactomorphic in every finite cover.","tokens_in":18306,"tokens_out":23092,"duration_ms":242455,"significance":"The main theorem gives the first hyperbolic L-spaces with arbitrarily many pairwise inequivalent pseudo-Anosov flows and no taut foliations, answering strengthened forms of questions of Calegari and of Min and Nonino. The construction is explicit and the distinguishing invariant is robust: the maximal prong count of a singular orbit is a conjugacy invariant, and the paper supplies precise numerical bounds for all competing prong counts. The appendix, proving that large rational surgeries on L-space links are L-spaces, is a useful and essentially self-contained contribution. The overall architecture — Fried surgery, FDTC computations, Birkhoff sections, and cylindrical contact homology — is coherent and the derived contact-theoretic consequences are significant if the main flow-theoretic claim holds. The principal weakness is that a single, localized computation of c_{B_0}(h) is the load-bearing step for the orbit inequivalence result, and the proof of that computation is not fully documented in the version under review.","major_comments":[{"comment":"The computation c_{B_0}(h) = -1/4 is load-bearing for Theorem 1.1. The proof passes to the quotient by the involution τ and asserts the chain of equivalences c_{B_0}(h) = -1/4 ⇔ c_{\\bar{B}_0}(\\bar{h}) = -1/2 ⇔ c_{\\bar{B}_0}(\\bar{h}^2) = -1 ⇔ c_{\\bar{B}_0}(\\bar{h}^2 ∘ D_δ) = 0, but the factor-2 relation between c_{B_0}(h) and c_{\\bar{B}_0}(\\bar{h}) is not justified, and the final verification that \\bar{h}^2 ∘ D_δ sends α to the right and β to the left at \\bar{B}_0 is only indicated through Figures 4–5, which are not reproduced in the text. Since Lemma 3.3 derives d_0 = ℓ_0(-6 μ_0 + λ_0) from this FDTC value, and Proposition 3.4 uses the resulting unique maximal prong count ℓ_0(M^{k+1} + 24) to prove orbit inequivalence, an error in this computation would invalidate the main theorem. Please provide a complete explicit verification — for instance the invariant train track computation mentioned in Remark 3.6 — or a detailed algebraic calculation of the images of α and β, and justify the stated equivalences.","section":"Sec. 3, Proposition 3.1"},{"comment":"The proof that the constructed flows remain orbit inequivalent in finite covers is asserted without an explicit justification. The paper states in the proof of Theorem 1.6 that the flows are distinguished by the maximal number of prongs and therefore their lifts to any finite cover are also orbit inequivalent. This relies on the fact that the prong count of a singular orbit is preserved under lifting to a finite cover, which is true but should be stated, since a singular orbit can have multiple lifts and a reader needs to know that each lift has the same number of prongs as its projection. Adding one sentence with this justification would make the finite-cover claim in Theorem 1.6 fully supported.","section":"Sec. 3, Proposition 3.4 and Theorem 1.6"}],"minor_comments":[{"comment":"There is a typo: the tuple in “-L_n(r_0, . . . , r_n)” should be “-L_n(r_0, . . . , r_{n-1})”.","section":"Sec. 3, proof of Theorem 1.6"},{"comment":"The distance notation Δ(d_i, r_i) := |d_i · r_i| would benefit from a brief explanation of the dot product convention for slopes, since d_i is sometimes a multicurve and sometimes a vector of intersection numbers.","section":"Sec. 2.3"},{"comment":"The sentence “each p_i is odd since it is coprime to q_i” is only true because every q_i is even; it would be clearer to say “each p_i is coprime to the even q_i, hence odd.”","section":"Lemma 3.7"},{"comment":"The abstract says “for each n ∈ N,” but Theorem 1.1 is stated for even n ≥ 4. The consequence for arbitrary n follows by taking a larger even N and selecting n of the N flows, but the phrasing could be clarified to avoid appearing to state Theorem 1.1 for all n.","section":"Abstract and Introduction"},{"comment":"The proof of Proposition 3.1 relies heavily on Figures 3–5, but these figures are not included in the version under review; even if the final submission contains them, the argument should indicate what feature of the figures establishes the right/left veering behavior of the arcs.","section":"Figures"}],"recommendation":"major_revision","confidential_remarks":"The central construction is explicit and the overall argument is convincing, but the FDTC computation in Proposition 3.1 is a localized, load-bearing gap. The authors should be asked to supply the missing computation before acceptance. The paper also depends on the unpublished preprint [Zun24] for the contact-form construction; the editors may wish to check whether that reference has been accepted or refereed, since the contact-theoretic conclusions rely on it. Apart from this, the manuscript is a strong fit for a top journal."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's the thing you should know: for each even n this paper constructs infinitely many hyperbolic L-spaces carrying n pairwise orbit-inequivalent pseudo-Anosov flows. That is a genuinely new result; previous many-flow examples were toroidal or had b1>0. The authors also use the flows to answer Calegari's question about quasigeodesic flows without taut foliations and to disprove Min-Nonino's conjecture about universally tight contact structures on hyperbolic L-spaces.\n\nWhat's good: the construction is concrete. They take the chain link L_n, use the fibered structure and a suspension pseudo-Anosov flow, then perform Fried surgery with carefully chosen quarter-integer slopes. The distinguishing invariant—maximal prong count at the surgery core—is simple and robust. The rotational symmetry gives n rotations of the same surgery, and the prong count changes because the degeneracy slope at one boundary component is not rotationally invariant. The appendix proving large rational surgeries on L-space links are L-spaces is also solid and independently useful.\n\nThe soft spot is exactly where you'd guess: Proposition 3.1, the claim that the FDTC of the monodromy at B0 is -1/4. The proof passes to a quotient by an involution and asserts a diagrammatic fact about the images of two arcs (Figures 4-5). The figures aren't reproduced, so the reader cannot check without the paper in hand. This computation is load-bearing: it determines the degeneracy slope d0, which makes the prong count at the core uniquely maximal. If c_B0(h) were 0, the argument for orbit inequivalence would collapse. That doesn't mean it's wrong—the claim looks plausible and Remark 3.6 says an invariant train track computation gives the finer data—but a referee should verify it independently. It's the one place where the paper asks you to take a picture on faith.\n\nThe rest of the logic checks out. No circularity: the cited results from Zung and Baldwin-Sivek are prior independent theorems, not the target. The contact-structure part uses cylindrical contact homology and Barthelme-Frankel-Mann's reconstruction theorem, which is appropriate.\n\nVerdict: this deserves a serious referee. The main theorem is important, the construction is explicit, and the potentially weak step is localized and checkable. I'd send it out and ask specifically for a careful verification of the FDTC computation. If that holds, the paper is a genuine advance.","headline":"First hyperbolic L-spaces with arbitrarily many orbit-inequivalent pseudo-Anosov flows; the construction is explicit and novel, with one diagrammatic FDTC computation as the main thing to check.","tokens_in":18817,"tokens_out":2181,"would_cite":true,"duration_ms":23937,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["57K10","57K32","57K33","57K18"],"pacs":[],"model":"deepseek-v4-flash","headline":"For every even n, infinitely many hyperbolic L-spaces carry n pairwise distinct pseudo-Anosov flows, and hence have no taut foliations.","keywords":["pseudo-Anosov flows","L-spaces","hyperbolic 3-manifolds","Dehn surgery","Fried surgery","fractional Dehn twist coefficient","Birkhoff sections","universally tight contact structures"],"falsifier":"Independently compute the fractional Dehn twist coefficient of the monodromy $h$ at $B_0$, for instance by constructing an invariant train track or conjugating the explicit Dehn-twist word into a braid, and check whether $c_{B_0}(h) = -1/4$; equivalently, check whether the maximal prong counts $\\ell_0(M^{k+1}+24)$ for $k=0,\\ldots,n-1$ are indeed distinct for the stated surgery slopes.","tokens_in":17844,"feed_emoji":"🌀","tokens_out":10893,"duration_ms":104854,"temperature":0.7,"pith_summary":"The paper proves that topological rigidity and dynamical richness can coexist on the same hyperbolic 3-manifold. For every even $n\\geq 4$ it constructs infinitely many hyperbolic L-spaces, obtained by quarter-integer Dehn surgery on an $n$-component chain link, each carrying $n$ pseudo-Anosov flows no two of which are orbit equivalent; because these manifolds are L-spaces with odd-order first homology, they admit no taut foliations. The flows have no perfect fits, so they are quasigeodesic, and they admit negative Birkhoff sections, so after reversing orientation they give $n$ universally tight contact structures that remain pairwise non-contactomorphic in every finite cover. The distinguishing dynamical invariant is concrete: the number of prongs at a distinguished closed orbit takes distinct values on rotated copies of the same surgery. These results answer more general versions of existing questions about quasigeodesic flows on manifolds without taut foliations and about universally tight contact structures on hyperbolic L-spaces.","feed_headline":"Hyperbolic L-spaces can hold arbitrarily many distinct flows","feed_subtitle":"Quarter-integer chain-link surgeries give infinitely many manifolds with n non-equivalent pseudo-Anosov flows and no taut foliations.","key_machinery":"The central mechanism is Fried surgery applied to the suspension flow of a fibered hyperbolic link, which extends a pseudo-Anosov flow on a link complement to a flow on the Dehn-surgered manifold by collapsing each boundary torus along the surgery slope; the collapsed core is then a closed orbit whose number of prongs equals the intersection distance $\\Delta(d_i,r_i)$ between the degeneracy slope and the surgery slope. The other load-bearing identity is the formula $d_i = q_i\\mu_i' + k_i\\lambda_i'$ relating the fractional Dehn twist coefficient of the fiber monodromy to the degeneracy slope, together with the computation $c_{B_0}(h) = -1/4$. The rotational symmetry of the chain link turns one surgery into $n$ homeomorphic ones, and the distinct maximal prong counts $\\ell_0(M^{k+1}+24)$ of the distinguished core orbit certify orbit inequivalence. Negative Birkhoff sections from the same fibration, after orientation reversal, feed the contact-geometric part of the argument.","core_discovery":"The central claim, Theorem 1.1, is that each even $n\\geq 4$ admits infinitely many hyperbolic L-spaces of the form $L_n(r_0,\\ldots,r_{n-1})$ with $r_i = M^{i+1}/4$ for odd $M>8n$, and each such manifold carries $n$ distinct pseudo-Anosov flows. The flows arise by Fried surgery from the suspension pseudo-Anosov flow of the fibered chain link $L_n'$, and the key computation is that the monodromy's fractional Dehn twist coefficient at the distinguished boundary component is $-1/4$ while it vanishes at the others. This yields degeneracy slopes $d_0 = \\ell_0(-6\\mu_0+\\lambda_0)$ and $d_i=\\ell_i\\mu_i$ for $i\\neq 0$. The rotational symmetry of the chain link gives $n$ homeomorphic copies of the same surgery, and on the $k$-th copy the core of the zeroth surgery torus has $\\ell_0(M^{k+1}+24)$ prongs; since these counts are distinct for $k=0,\\ldots,n-1$, the corresponding flows are pairwise orbit inequivalent. Long surgeries are hyperbolic L-spaces by standard results, and odd-order first homology rules out taut foliations.","pith_inferences":["The same construction may work for other fibered hyperbolic links whose monodromy has a nonzero fractional Dehn twist coefficient at exactly one boundary component and whose symmetry group orbits that component; the chain link is one instance of a general recipe.","Because the flows are distinguished by maximal prong count, a coarser invariant than the full orbit set, the rotated copies might admit further inequivalent flows or additional distinctions that the paper does not pursue.","The paper's Theorem 1.1 is stated for even $n\\geq 4$, and the proof uses an involution that pairs boundary components; extending the computation to odd $n$ or to larger symmetry groups would likely remove this restriction.","The contact-structure conclusion suggests that hyperbolic L-spaces, which have no taut foliations, are nevertheless natural hosts for universally tight contact structures, so the dichotomy between foliations and contact geometry on such manifolds is less stark than it might appear."],"forward_implications":["For each even $n\\geq 4$ there are infinitely many hyperbolic L-spaces with $n$ pairwise orbit-inequivalent pseudo-Anosov flows; in particular these manifolds admit no taut foliations.","Since these flows have no perfect fits, they are quasigeodesic, so these are hyperbolic manifolds without taut foliations that still carry quasigeodesic flows.","The orientation-reversed manifolds carry $n$ universally tight contact structures whose lifts to any finite cover are pairwise non-contactomorphic.","Large rational surgeries on any hyperbolic L-space knot yield a universally tight contact structure and a quasigeodesic pseudo-Anosov flow.","An atoroidal rational homology 3-sphere has only finitely many pseudo-Anosov flows with positive Birkhoff sections, up to orbit equivalence by an isotopy."],"supporting_citations":[{"why":"Establishes hyperbolicity of the $n$-chain link, the base link used for all surgeries.","marker":"[NR92]"},{"why":"Proves the chain link is an L-space link, which is needed for the surgered manifolds to be L-spaces.","marker":"[Liu17]"},{"why":"Provides the universal bounds showing Dehn surgeries on the hyperbolic chain link are hyperbolic for long slopes.","marker":"[HK05]"},{"why":"Supplies the theorem that pseudo-Anosov flows without perfect fits are quasigeodesic, and that suspension flows have no perfect fits.","marker":"[Fen12]"},{"why":"Defines fractional Dehn twist coefficients and gives the degeneracy-slope formula used throughout.","marker":"[HKM08]"},{"why":"Gives the smooth model for transitive topological Anosov flows, used to upgrade Fried-surgery flows to genuine pseudo-Anosov flows.","marker":"[Sha21]"},{"why":"Shows a pseudo-Anosov flow is determined up to isotopy orbit equivalence by its closed orbits, the bridge to contact homology.","marker":"[BFM25]"},{"why":"Constructs stable Hamiltonian and contact forms with prescribed Reeb dynamics from positive Birkhoff sections.","marker":"[Zun24]"},{"why":"Provides the inequality used to show the constructed manifolds have first-homology orders growing to infinity, hence infinitely many distinct manifolds.","marker":"[Ost37]"}],"fun_headline_variants":["Even n≥4: hyperbolic L-spaces with n distinct pseudo-Anosov flows","Infinitely many hyperbolic L-spaces with n non-equivalent flows for each even n","n distinct quasigeodesic flows without perfect fits on hyperbolic L-spaces","Chain-link surgeries give n orbit-inequivalent pseudo-Anosov flows","Hyperbolic L-spaces realize arbitrarily many distinct flows"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing computation is that the monodromy of the fibered chain link has fractional Dehn twist coefficient exactly $-1/4$ at the distinguished boundary component; this value is established by passing to a quotient by an involution and reading arc images from diagrams, and if it were wrong the degeneracy slope, the prong counts, and the orbit inequivalence would not follow.","fun_headline_variants_meta":{"raw":{"variants":["Even n≥4: hyperbolic L-spaces with n distinct pseudo-Anosov flows","Infinitely many hyperbolic L-spaces with n non-equivalent flows for each even n","n distinct quasigeodesic flows without perfect fits on hyperbolic L-spaces","Chain-link surgeries give n orbit-inequivalent pseudo-Anosov flows","Hyperbolic L-spaces realize arbitrarily many distinct flows"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000553,"raw_usage":{"total_tokens":2630,"prompt_tokens":935,"completion_tokens":1695,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":551,"completion_tokens_details":{"reasoning_tokens":1595}},"tokens_in":551,"tokens_out":1695,"duration_ms":13576,"temperature":1.0,"reasoning_tokens":1595,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T13:36:46.225034+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Independently compute the fractional Dehn twist coefficient of the monodromy $h$ at $B_0$, for instance by constructing an invariant train track or conjugating the explicit Dehn-twist word into a braid, and check whether $c_{B_0}(h) = -1/4$; equivalently, check whether the maximal prong counts $\\ell_0(M^{k+1}+24)$ for $k=0,\\ldots,n-1$ are indeed distinct for the stated surgery slopes.","supporting_citations":[],"review_version":1}