{"id":"77f46bde-e963-4a59-b0e5-755d2eae26f8","arxiv_id":"1908.09079","paper_version":4,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For generic volume-preserving diffeomorphisms, a minimal expanding foliation with a matching-index periodic point is stably minimal, yielding robust mixing and an essentially dense hyperbolic ergodic component.","lead":"Many volume-preserving dynamical systems have invariant expanding leaf structures called foliations. The paper proves that, for typical such systems, a minimal expanding foliation with a matching periodic point stays minimal under small perturbations, forcing topological mixing and a dense hyperbolic ergodic component.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem B rests on an unproved assertion that every leaf of the minimal expanding foliation contains a well-placed k-strip in the superblender ball; minimality alone does not yield this, so the core mechanism is unsupported.","rationale":"The reader's verdict is CONDITIONAL, and I agree that the most load-bearing unresolved point is the geometric transfer in Section 3: minimality of W_f plus a matching periodic point is used to claim that every leaf contains a well-placed k-strip in the fixed superblender ball, and that this persists for W_g. This claim is the only bridge between the abstract minimality hypothesis and the superblender property, and without it Theorem B does not follow. Minimality alone does not imply the presence of large, flat, well-aligned discs in every leaf; a dense leaf can be highly folded and avoid a prescribed direction on a prescribed ball. The proof would need an additional argument, for example using the expanding property of the foliation and some form of recurrence to pull large strips back into B. The persistence of W_g itself is also asserted via a dominated splitting TM=TW_f⊕F, which is not established by Theorem 3.1; this is especially delicate in Case 2 where u(p)<u(q). These are not mere cosmetic omissions because they are used to conclude the essential density of Phc_g(q_g) for every x∈M. The reader's second flagged premise, Phc^W(p)=M, is actually derivable from minimality: the set Phc^W(p) is open and W-saturated, contains p because the stable and unstable spaces at p are complementary, and hence equals M by minimality. This part of the reader's concern is repairable. Since the main theorem may still be true with additional arguments and the proof relies on deep cited machinery, I do not move the verdict to REJECT; the appropriate disposition remains CONDITIONAL, requiring the authors to supply the missing geometric justification or a counterexample.","tokens_in":13022,"tokens_out":24337,"duration_ms":241850,"concrete_test":"Check whether the set A = {x∈M : the W-leaf through x contains a well-placed k-strip in B^ls_Λ(x0)} is actually equal to M under the stated hypotheses. In particular, verify whether A is f-invariant; if not, minimality alone cannot imply A=M because minimality propagates only invariant, open, saturated sets. A concrete model to test the local claim is T^3 with dominated splitting E^uu⊕E^u⊕E^s and rates (3,2,1/3), taking W tangent to the middle direction E^u and k=1; if a W-leaf can pass through the blender ball without containing a large disc C^1-close to E^uu, the assertion fails. If instead the assertion is proved in [ACW17, Corollary D] or [NH20, Lemma 3.2], locate the exact statement and adapt it; otherwise the proof of Theorem B has a genuine gap.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The hinge of Theorem B is the sentence in Section 3: 'Since W_f is minimal, all leaves of W_f contain a well-placed unstable k-strip in B^ls_Λ(x0), where k = dim W_f = u(p). The neighborhood U can be chosen so that all leaves of W_g also contain a well-placed unstable k-strip.' This is asserted without proof, and it is not a formal consequence of minimality. Minimality gives density of leaves, but a dense submanifold need not contain a geodesic k-disc of radius much larger than the blender ball, centered inside B, and C^1-close to the specific sum E^u_1⊕...⊕E^u_k from the ACW17 horseshoe splitting. The foliation W is only an expanding foliation of dimension k; its tangent space can be a different k-dimensional invariant subbundle of E^+, e.g. the slower unstable bundle E^u in a three-way dominated splitting, which is not close to the strongest unstable direction E^u_1. The property 'the leaf contains a well-placed strip in the fixed ball B' is W-saturated and open, but it is not f-invariant, so the standard minimality-open-saturated argument does not force it to hold for every leaf. The companion claim that generically TW_f has a dominated complement TM=TW_f⊕F, which is needed for the continuation W_g to exist, is likewise asserted without proof and is not a consequence of Theorem 3.1 as stated, especially in the case u(p)<u(q) treated later. By contrast, the reader's secondary concern that Phc^W(p)=M is never derived can be repaired: Phc^W(p) is open, W-saturated, and contains p because E^u(p)⊕E^s(p)=T_pM, so minimality forces it to be all of M.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies C1-generic volume-preserving diffeomorphisms and asks when a minimal expanding invariant foliation is stably minimal. Theorem A claims that, generically, if W is a minimal expanding foliation of dimension u and there is a hyperbolic periodic point p of unstable index u(p)=dim W, then W is stably minimal; in particular all C2 volume-preserving diffeomorphisms in a C1 neighborhood are topologically mixing. Theorem B claims that, under the same hypotheses, every such C2 diffeomorphism has a hyperbolic ergodic component whose essential closure is all of M, and the component is Bernoulli. The proof of Theorem B uses a superblender created from the horseshoe produced by the Avila–Crovisier–Wilkinson theory, together with Pesin homoclinic classes and a criterion for ergodicity. Theorem A is then derived from Theorem B, a minimality criterion, and a chain of blenders when the unstable index of p is strictly lower than that of the generic periodic point q. Section 5 proposes examples, including non-partially-hyperbolic ones, via a criterion for approximating by diffeomorphisms with stably minimal foliations.","tokens_in":13331,"tokens_out":16359,"duration_ms":159380,"significance":"If the main results are correct, they provide a new mechanism for stable minimality, stable topological mixing, and essentially dense hyperbolic ergodic components that does not require partial hyperbolicity. This would substantially extend earlier work by Bonatti–Díaz–Ures, Pujals–Sambarino, and the authors' own dimension-three results, and it would supply new non-partially-hyperbolic examples. The paper draws on deep external results and has no fitted parameters or definitional circularity. However, the proof contains several load-bearing assertions that are not justified in the text, so the significance can only be assessed after those gaps are closed.","major_comments":[{"comment":"The assertion that minimality of W_f implies that every leaf contains a well-placed unstable k-strip in the fixed superblender ball B^ls_Λ(x0) is not proved and is not a formal consequence of minimality. Minimality gives that each leaf is dense, but a dense k-dimensional immersed submanifold need not contain a large k-disc centered inside B^ls_Λ(x0) that is C1-close to the specific subbundle E^u_1⊕...⊕E^u_k of the horseshoe Λ. The well-placed strip property is open and W-saturated, but it is not f-invariant, so the standard minimality-open-saturated argument does not force it to hold on every leaf. This step is load-bearing: it is the only mechanism in Theorem B that connects an arbitrary W_g-leaf to the stable manifolds of Λ_g through the superblender property.","section":"§3, paragraph after Theorem 3.4"},{"comment":"The claim that, generically, the existence of an expanding invariant foliation W_f implies a dominated splitting TM=TW_f⊕F is asserted without proof or reference. Theorem 3.1 provides a dominated splitting of the zipped Oseledets splitting E^+⊕E^-, but if TW_f is a proper subbundle of E^+, it does not automatically follow that TW_f has a dominated complement. This assertion is needed for the very definition of stable minimality, namely for the existence and continuity of the continuation W_g for all g in a C1 neighborhood, and also for quasi-transversality between W_g and stable manifolds. A proof or a precise citation is required.","section":"§3, paragraph after Theorem 3.1"},{"comment":"In the proof of Theorem A, Case u(p)<u(q), the text states 'Since Phc^W(p)=M' and then applies Lemma 4.2, but this equality is never derived. The hypotheses of Theorem A include only minimality of W and the existence of p with u(p)=dim W; neither minimality alone nor the later argument showing that W^u(q_g) is dense implies that every leaf W(x) quasi-transversely intersects W^s(o(p)). Without Phc^W(p)=M, Proposition 4.1 cannot be applied, and the proof of Theorem A in this case collapses. The authors need either to prove this equality or to modify the argument.","section":"§4, Case 2"},{"comment":"The abstract promises that the hyperbolic ergodic component is Bernoulli and that all C2 diffeomorphisms in the neighborhood are topologically mixing, but the proof does not establish either property. The proof of Theorem B ends with the conclusion that x∈Phc_g(q_g)^ess for every x, and no argument for topological mixing appears in the proof of Theorem A. The Bernoulli assertion is not mentioned in the proof of Theorem B at all. Relatedly, Lemma 3.5 states W^s(q_g)=Phc_g(q_g)^ess and W^u(q_g)=Phc_g(q_g)^ess, but the proof only yields an inclusion after taking essential closures; as sets this equality is impossible when Phc_g(q_g)^ess=M and q_g has nontrivial stable and unstable index. The notation must be clarified and the properties actually needed in Lemmas 3.6 and the final step of Theorem B must be stated and proved.","section":"Abstract and Lemmas 3.5–3.6"}],"minor_comments":[{"comment":"The notation for the superblender ball is inconsistent (Bls, B^ls, B^ls_Λ), and the definition of 'well-placed unstable k-strip' would benefit from a precise statement about the size of the radius relative to the ball and the allowed C1 distance to the subbundle.","section":"§3, Definition 3.2"},{"comment":"The statement of Theorem 4.6 contains a typo: 'indices u and (u+c1)' should presumably be 'u and u+c' or similar, and the following sentence is grammatically incomplete.","section":"§4, Theorem 4.6"},{"comment":"The equality W^s_f(p)=W^s_f0(p) is written as a set equality; since these are leaves of a foliation, the notation should specify whether equality means equality as leaves or equality of their closures, and the proof of the internal-radius estimate could be expanded.","section":"§5.1.1"}],"recommendation":"major_revision","confidential_remarks":"The gaps identified in the major comments are substantial, especially the well-placed strip assertion in Section 3. If the authors cannot supply a proof or a reference for that assertion, the main theorem of the paper is not established. I would encourage the editors to treat this as a true conditional acceptance: the central idea is attractive, but the missing steps are load-bearing rather than cosmetic."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague — this is a real step forward in the Pugh–Shub program. Theorems A and B remove the partial hyperbolicity requirement from stable minimality of expanding foliations, and Section 5 gives explicit non-partially-hyperbolic examples. The core idea — a minimal expanding foliation plus a matching hyperbolic periodic point gives stable minimality — is new and is exactly the right kind of mechanism. If the proof works, it is an important result.\n\nThe paper is honest about its debts: it builds on ACW17, HHTU11, AB12, and the authors' own NH20. The statements are clear, and the Minimality Criterion (Prop 4.1) is clean and independently useful. The overall strategy is understandable: use generic dominated splitting to get a continuation, use a superblender to force every W_g leaf to quasi-transversely intersect the stable lamination of a horseshoe, then feed that into the HHTU ergodicity criterion.\n\nThe soft spot is the load-bearing sentence in §3: minimality of W_f is claimed to imply every leaf contains a well-placed unstable k-strip in the fixed superblender ball, and that this persists under perturbation. That is not a formal consequence of minimality. A dense leaf need not contain a large disc of a specific tangent direction inside a fixed ball. The property is W-saturated and open, but not f-invariant, so the standard minimality-open-saturated argument does not force it. The stress-test note is on target here. Also asserted rather than proved is the claim that generically TW_f has a dominated complement, so W_g continues. Theorem 3.1 gives TM=E+⊕E-, not directly the TW_f⊕F splitting needed, and the u(p)<u(q) case makes that harder. These are the two real gaps. The secondary concern about Phc^W(p)=M in §4 is repairable: that set is open, W-saturated, and contains p, so minimality does force it to be all of M.\n\nThis does not sink the paper. The gaps are likely repairable with additional arguments, but as written the main theorem is not fully proven. The reader's CONDITIONAL verdict is the right one. I would send it to a serious referee, not desk-reject, because the result is important and the strategy is plausible. I would not cite the theorem as proved yet, but I would want to see the revised version. Recommend refereeing, with a request to fill the strip lemma and the continuation argument.","headline":"Significant extension of stable minimality beyond partial hyperbolicity, but the proof's main mechanism relies on an unproved geometric assertion that does not follow from minimality.","tokens_in":13925,"tokens_out":2992,"would_cite":false,"duration_ms":29710,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["37C05","37C40","37C85","37D25","37D30"],"pacs":[],"model":"deepseek-v4-flash","headline":"A minimal expanding foliation plus one matching hyperbolic periodic point is stably minimal for generic volume-preserving diffeomorphisms.","keywords":["stable minimality","expanding foliation","minimal foliation","hyperbolic periodic point","superblender","Pesin homoclinic class","topological mixing","stable ergodicity"],"falsifier":"Exhibit a $C^1$-generic volume-preserving diffeomorphism $f$ with a minimal expanding foliation $W$ and a hyperbolic periodic point $p$ with $u(p)=\\dim W$, and a $C^2$ map $g$ arbitrarily $C^1$-close to $f$ whose continuation $W_g$ has a non-dense leaf; equivalently, find a $W_f$-leaf that contains no well-placed unstable $k$-strip inside the superblender ball $B^{\\mathrm{ls}}_\\Lambda(x_0)$ produced by Theorem 3.3, since such a leaf would break the quasi-transverse intersection step and the essential density of the ergodic component.","tokens_in":12776,"feed_emoji":"🌀","tokens_out":19260,"duration_ms":163729,"temperature":0.7,"pith_summary":"This paper claims a robust mechanism for stable minimality in volume-preserving dynamics. For a $C^1$-generic diffeomorphism preserving a smooth volume, if an expanding invariant foliation $W$ is minimal and a hyperbolic periodic point has unstable index equal to $\\dim W$, then $W$ is stably minimal: its continuation remains minimal for every $C^2$ volume-preserving diffeomorphism in a $C^1$-neighborhood. If true, the same condition makes all such nearby $C^2$ maps topologically mixing, and each carries a hyperbolic ergodic component whose essential closure — the points every neighborhood of which meets the component with positive measure — is the whole manifold, and this component is Bernoulli. The proof turns minimality into a geometric intersection statement inside superblenders, and the final section builds examples that are not partially hyperbolic.","feed_headline":"One matching periodic point makes a minimal foliation stably minimal","feed_subtitle":"For C1-generic volume-preserving maps, nearby C2 systems stay topologically mixing and Bernoulli.","key_machinery":"The central mechanism is the s-stable superblender: an open ball around a point of a horseshoe $\\Lambda$ such that every well-placed unstable $k$-strip — a $k$-disc of radius much larger than the ball, almost tangent to the expanding subbundle — quasi-transversely intersects the stable manifold of some point of $\\Lambda$, with the property holding $C^1$-robustly. Minimality of the expanding foliation is used to ensure every leaf contains such a strip, so every $W_g$-leaf meets $W^s(\\Lambda_g)$ quasi-transversely, meaning the tangent spaces share no nonzero vector. Those intersections place every point inside the essential closure of the Pesin homoclinic class $\\mathrm{Phc}^g(q_g)$, the set of points whose stable Pesin manifold meets the unstable manifold of $q_g$; the ergodicity criterion of [HHTU11] turns this class into a hyperbolic ergodic component. In the harder case $u(p)<u(q)$, the complementary machinery is a chain of u-blenders — open sets in which every well-placed disc of the appropriate dimension intersects the unstable manifold of a lower-index periodic point — forcing the inclusion $W^u(q_g)\\subset\\cdots\\subset W^u(p_g)$. The Minimality Criterion (Proposition 4.1) is the final switch: dense $W^u(p_g)$ together with $\\mathrm{Phc}^{W_g}(p_g)=M$ implies $W_g$ is minimal.","core_discovery":"The paper's central claim is Theorem A: for a residual subset of $\\mathrm{Diff}^1_m(M)$, whenever $W$ is a minimal expanding $f$-invariant foliation and there is a hyperbolic periodic point $p$ with unstable index $u(p)=\\dim W$, the foliation is stably minimal — there exists a $C^1$-neighborhood $\\mathcal{U}$ of $f$ such that every $g\\in\\mathcal{U}\\cap \\mathrm{Diff}^2_m(M)$ has a minimal $g$-invariant continuation $W_g$. In particular every such $g$ is topologically mixing. Theorem B isolates the dynamical content of the mechanism: for each such $g$, the Pesin homoclinic class $\\mathrm{Phc}^g(q_g)$ of the periodic point supplied by the generic positive-entropy dichotomy is a hyperbolic ergodic component whose essential closure is $M$, and the component is Bernoulli. The superblender does the work: minimality of $W_f$ guarantees that every leaf contains a well-placed unstable strip inside a superblender ball, so every $W_g$-leaf quasi-transversely meets the stable manifolds of a horseshoe; this places every point of $M$ inside the essential closure of $\\mathrm{Phc}^g(q_g)$. When the given periodic point $p$ has smaller unstable index than $q$, a chain of $u$-blenders transfers the density of $W^u(q_g)$ down to $W^u(p_g)$, and a minimality criterion converts dense unstable manifold plus $\\mathrm{Phc}^{W_g}(p_g)=M$ into minimality of $W_g$.","pith_inferences":["A natural sharpness test is to drop the periodic-point hypothesis: if a minimal expanding foliation with no hyperbolic periodic point of matching unstable index can fail to be stably minimal, then the index condition in Theorem A is necessary rather than a technical convenience.","The proof uses the volume-preserving assumption at superblender creation and at the measure estimate for the Pesin class; transferring the argument to dissipative diffeomorphisms would require new tools at exactly those two steps, so those are the places to look for a counterexample or an extension.","One quantitative prediction of the mechanism is that the radius of the $C^1$-neighborhood of stable minimality is controlled by the superblender's size and by how uniformly minimal leaves supply well-placed strips; computing these quantities in the Section 5 examples would give concrete lower bounds for the neighborhood.","The unproved assertion that $\\mathrm{Phc}^W(p)=M$ in Section 4 could be verified in the Section 5 examples, where every leaf of the expanding foliation cuts the dense stable manifold of the fixed point; closing that gap would remove the main missing step between Theorem B and Theorem A in the case of equal indices."],"forward_implications":["If Theorem A is correct, then for $C^1$-generic volume-preserving diffeomorphisms a minimal expanding foliation of dimension $u$ is stably minimal as soon as some hyperbolic periodic point of unstable index $u$ exists: the continuation remains minimal for every $C^2$ map in a $C^1$-neighborhood.","Every $C^2$ map in the $C^1$-neighborhood is topologically mixing, so minimal-foliation-plus-one-periodic-point is a mechanism for robust mixing without partial hyperbolicity.","Theorem B shows that the same hypotheses give, for all nearby $C^2$ maps, an explicitly defined hyperbolic ergodic component whose essential closure is the whole manifold and whose dynamics is Bernoulli.","The Section 5 examples show that stably minimal expanding foliations occur among diffeomorphisms that are not partially hyperbolic, so the mechanism is not a corollary of a dominated center splitting.","Proposition 5.1 supplies a checkable sufficient condition — every leaf intersects an open set that meets the local stable manifold of a periodic point — for residual minimality of the continuation and, by Theorem A, for dense stable minimality in a $C^1$-neighborhood."],"supporting_citations":[{"why":"Provides the generic dichotomy used as Theorem 3.1: positive Lyapunov exponents force ergodicity and a dominated zipped Oseledets splitting with a dense Pesin class.","marker":"[Mn84]"},{"why":"Contributes the zero-Lyapunov-genericity half of Theorem 3.1.","marker":"[Boc02]"},{"why":"Supplies Lemma 3.7, the saturation of invariant sets by stable and unstable manifolds, and part of the positive-entropy dichotomy.","marker":"[Her12]"},{"why":"Establishes the generic positive-metric-entropy dichotomy that Theorem 3.1 rests on.","marker":"[ACW16]"},{"why":"Creates the s-stable superblender (Theorem 3.3), the geometric engine that turns leaf intersections into essential density.","marker":"[ACW17]"},{"why":"Gives the generic homoclinic relatedness of same-index periodic points (Theorem 3.4) used to connect p, q, and the horseshoe.","marker":"[AC12]"},{"why":"Supplies the ergodicity criterion (Theorem 3.9) that makes the Pesin homoclinic class a hyperbolic ergodic component.","marker":"[HHTU11]"},{"why":"Creates u-blenders (Theorem 4.4), the mechanism that transfers unstable-manifold density from higher to lower index.","marker":"[HHTU10]"},{"why":"Provides Lemma 3.8 and the measure estimate showing the Pesin class of q has measure arbitrarily close to 1 for nearby C2 maps.","marker":"[AB12]"},{"why":"Supplies Lemma 4.2 on the C1-persistence of Phc^W(p)=M and the dimension-3 minimality result being extended here.","marker":"[NH20]"}],"fun_headline_variants":["One periodic point makes minimal foliations stably minimal","Stable minimality triggered by a periodic point","Generic stable minimality from matching unstable index","One periodic point yields stable minimality in C1 neighborhood","Minimal expanding foliation becomes stably minimal generically"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is the unstated geometric transfer that minimality of $W_f$ forces every $W_f$-leaf to contain a well-placed unstable $k$-strip inside the fixed superblender ball $B^{\\mathrm{ls}}_\\Lambda(x_0)$, with a $C^1$-robust continuation to $W_g$, together with the unproved assertion 'Since $\\mathrm{Phc}^{W}(p)=M$' in Section 4; if either of these gives way, the proof collapses.","fun_headline_variants_meta":{"raw":{"variants":["One periodic point makes minimal foliations stably minimal","Stable minimality triggered by a periodic point","Generic stable minimality from matching unstable index","One periodic point yields stable minimality in C1 neighborhood","Minimal expanding foliation becomes stably minimal generically"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000646,"raw_usage":{"total_tokens":3011,"prompt_tokens":1032,"completion_tokens":1979,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":648,"completion_tokens_details":{"reasoning_tokens":1904}},"tokens_in":648,"tokens_out":1979,"duration_ms":13531,"temperature":1.0,"reasoning_tokens":1904,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:23:59.072982+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Exhibit a $C^1$-generic volume-preserving diffeomorphism $f$ with a minimal expanding foliation $W$ and a hyperbolic periodic point $p$ with $u(p)=\\dim W$, and a $C^2$ map $g$ arbitrarily $C^1$-close to $f$ whose continuation $W_g$ has a non-dense leaf; equivalently, find a $W_f$-leaf that contains no well-placed unstable $k$-strip inside the superblender ball $B^{\\mathrm{ls}}_\\Lambda(x_0)$ produced by Theorem 3.3, since such a leaf would break the quasi-transverse intersection step and the essential density of the ergodic component.","supporting_citations":[],"review_version":1}