{"id":"d0910cef-398b-47f0-9c3f-bf840c00c2d4","arxiv_id":"1908.05746","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"A non-annular torus homeomorphism with small wandering domains admits an irrational circle factor exactly when it has uniformly bounded rotational deviations in some rational direction.","lead":"This paper gives a complete answer to when a periodic point free map of a two-torus can be projected onto an irrational circle rotation. It shows that, outside of two special obstructions, a bounded twist in a rational direction guarantees such a projection.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 6.1's empty-interior construction requires boundary points of wandering blocks to lie in the open set T, but this membership is not established and is load-bearing.","rationale":"The Reader's verdict accepts the paper with moderate confidence and identifies the small-wandering-domains hypothesis as the weakest assumption. I agree that this hypothesis is the relevant stress point, but I locate the problem more precisely: it is not the hypothesis itself, but the way Lemma 6.1 uses it in the empty-interior case. The main theorem is a long structured argument, and most of the machinery — the ρ-centralized skew-product, Propositions 5.3–5.4, Theorem 5.5, and the use of Theorem 5.6 to produce the continua C^s — appears coherent. However, Lemma 6.1 is essential, and the membership assertion (71) seems to conflate the closure of a wandering block with the open set T. The proof needs the selected boundary points to lie in T and then in T \\ Γ^{-s}(T) for the height-decrease argument; neither inclusion is proved. This is not a disagreement with the theorem's truth, and it is not an ad hominem or a style objection; it is a precise missing step in a central lemma. A concrete re-derivation on a Denjoy-type example, or a formalization that type-checks the membership of w_1 in D, would settle whether the gap is real or merely a compressed argument. Until then, the proof of Theorem A should be regarded as conditional on repairing this step.","tokens_in":32361,"tokens_out":23435,"duration_ms":253606,"concrete_test":"Test Lemma 6.1 in the simplest empty-interior model: take one of the Denjoy-type pseudo-rotations from §3.2/§3.3 and compute T from (68)–(69) explicitly near a wandering block \\check W_n. Verify whether there exists u ∈ (−1/2, 1/2) such that Γ^{r_n+u}(0,\\hat z_n) ∉ T; if such a u exists, assertion (71) fails and the recursion producing the contradiction is invalid. If no such u exists, the missing proof that the Γ-orbit of ∂\\hat W_n is contained in T must be supplied from (69) and Proposition 5.3.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The proof of Theorem A ultimately rests on Lemma 6.1: it produces the F-invariant bounded open set T whose Γ-translates give the annular continua C^s used to build the semi-conjugacy. In the hard case where Ω(f) has empty interior, T is defined via (68)–(69) as U_F(\\hat V_{1/2,0}), with \\hat V containing B_1(\\hat z_0) and lifts of wandering domains meeting it, but not their boundaries. The argument then chooses \\hat z_n ∈ ∂\\hat W_n and asserts in (71) that Γ^{r_n+u}(0,\\hat z_n) belongs to \\overline{\\check W_n} ∩ T ∩ Ω(F) for every u ∈ (−1/2, 1/2). This membership in T is not a consequence of (69)–(70): T is open and is only F-invariant, not Γ-invariant, and ∂\\hat W_n is included in \\hat V only where it happens to lie in B_1(\\hat z_0). The subsequent construction of w_1 and the inductive sequence (73)–(74) requires these boundary points to lie in the open set D := T \\ Γ^{-s}(T); if they lie only on ∂T, the height estimate (75) does not follow. Since Lemma 6.1 is the only source of the separating continua used to construct the semiconjugacy, this is a load-bearing gap in the proof as written.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper characterizes periodic-point-free orientation-preserving homeomorphisms of the 2-torus with small wandering domains that admit an irrational circle rotation as a topological factor. Theorem A states that such a homeomorphism admits such a factor if and only if it has uniformly bounded rotational deviations in some rational direction. The proof introduces a ρ-centralized skew-product F on T×A, studies its recurrent set and bounded invariant open sets, and uses these to construct a family of essential annular continua whose boundaries define the semiconjugacy. Section 3 presents examples showing that the hypotheses in Theorem A are sharp; Corollaries 1.1 and 1.2 extend previous results of Jäger and characterize the Dehn-twist case.","tokens_in":32683,"tokens_out":21418,"duration_ms":194972,"significance":"If correct, Theorem A is a substantial result: it identifies the only obstructions (annularity and the geometry of wandering domains) to the existence of irrational circle factors for periodic-point-free torus homeomorphisms, extending work of Jäger and Jäger–Tal. The ρ-centralized skew-product machinery is an attractive and potentially reusable tool, and the Section 3 examples usefully delineate the sharpness of the 'small wandering domains' hypothesis. The paper is carefully structured and builds on standard external results (Brouwer's translation theorem, results of Jäger–Tal, Hauser–Jäger, Jäger–Passeggi). However, the proof of the central construction contains two load-bearing gaps that must be repaired before the main theorem can be regarded as established.","major_comments":[{"comment":"The proof asserts that for each wandering block there is a boundary point z_hat_n ∈ ∂W_hat_n such that Γ^{r_n+u}(0, z_hat_n) ∈ closure(W_check_n) ∩ T ∩ Ω(F) for |u| < 1/2. The membership in T is not established. The set V_hat in (69) is open and contains only the wandering domains themselves, not their boundaries; T = U_F(V_hat_{1/2,0}) is the open connected component of the union of F-iterates of the block, and there is no reason that the Γ-orbit of a boundary point of a wandering domain lies in that open set. The subsequent construction of the sequence (w_k) in (73)–(74) and the height estimate (75) require these points to lie in the open set D = T \\ Γ^{-s}(T). If the points lie only on ∂T, the inequality (75) does not follow, and the contradiction with the boundedness of D collapses. Since Lemma 6.1 is the only source of the separating continua C^s used to define the semiconjugacy, this is a load-bearing gap.","section":"§6.1, Lemma 6.1 (empty-interior case), Eqs. (69)–(71)"},{"comment":"The sentence 'Since f is not eventually annular, this implies the rotation set of any lift of f is singleton containing a totally irrational vector' is false. For example, the translation by (1/2, α) with α irrational is periodic point free, not eventually annular, and exhibits uniformly bounded rotational deviations in every direction, yet its rotation set is {(1/2, α)}, which is not totally irrational. The proof then invokes Lemma 5.7, whose hypothesis requires total irrationality, so the conclusion that U_F(V)_t is annular does not follow for such f. This is not a corner case: translations of this kind fall under the hypotheses of Theorem A, and the argument in §6.1 relies on Theorem 5.6 to obtain the two unbounded components of A \\ Γ^{-s}(T)_t. Without an additional argument for non-totally-irrational pseudo-rotations, Theorem 5.6 and consequently the construction of the continua C^s are unproved.","section":"§5, proof of Theorem 5.6, first case"}],"minor_comments":[{"comment":"The manuscript contains numerous OCR-style typos ('irra tional', 'W e', 'T ori', 'A TION', 'FR EE', 'F ACTORS'). The paper should be carefully proofread before resubmission.","section":"Abstract and title"},{"comment":"The definition of the product semiconjugacy H is written as H(x,y) = (h_1(x), h_2(y)), but the maps h_1, h_2 are defined on T^2, not on T. It should read H(z) = (h_1(z), h_2(z)) for z ∈ T^2.","section":"§4, Proof of Corollary 1.1"},{"comment":"The sentence 'by Proposition 5.1 we know that T ⊂ Ω(F)' is imprecise: the inclusion follows from the Γ-invariance and F-invariance of Ω(F) together with Theorem 5.5, not from Proposition 5.1 alone. The citation should be adjusted.","section":"§6.1, nonempty interior case of Lemma 6.1"}],"recommendation":"major_revision","confidential_remarks":"The two major comments concern the central proof of Theorem A. The first (Lemma 6.1, empty-interior case) is a precise technical gap in the construction of the separating continua; the second (Theorem 5.6) is a false implication that leaves an entire class of pseudo-rotations untreated. Both appear fixable in principle, but the fixes are nontrivial and affect the main theorem, so a major revision is appropriate. The paper's overall approach and the Section 3 examples are valuable, and I would be willing to consider a revised version."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is a strong, important paper, but I would not accept it as it stands. The main theorem genuinely extends Jäger–Tal: periodic point free, non-eventually annular, small wandering domains, and bounded rational-slope vertical deviations are together equivalent to having an irrational circle factor. The adaptation of the rho-centralized skew-product to Dehn twist isotopy classes is new, and the three Section 3 counterexamples do exactly the sharpness work. If the main proof is sound, this is a complete answer to a good question. The reader's report is fair on those points.\n\nThe soft spot is Lemma 6.1, specifically the empty-interior case. The proof needs, for each wandering block W_n, a boundary point z_n in ∂W_n whose Γ-line lies in T ∩ Ω(F), and then uses those points to push a sequence w_k inside the open sets D_n = T \\ Γ^{-s}(T). I think the stress-test note is right: T = U_F(V-hat_{1/2,0}) is open and F-invariant, not Γ-invariant, and the definition of V-hat includes the wandering domains themselves, not their boundaries. Equation (71) asserts Γ^{r_n+u}(0,z_n) ∈ closure(W_n) ∩ T ∩ Ω(F); the Ω(F) part follows from Theorem 5.5 since ∂W_n ⊂ Ω(f), but the T part does not follow from (69)–(70). And the later inequality (75) needs w_k ∈ D_k, i.e. a point in the open block over W_k, so a boundary point cannot be substituted. This is not a minor typo: it is the step that produces the F-invariant open set whose Γ-translates give the annular continua, and without it the semiconjugacy construction has no input. If the author can prove the Γ-lines of those boundary points actually lie in T, fine; as written I do not see which prior statement supplies that.\n\nThere is a smaller point: the proof leans on several substantial external theorems and on the author's own [KPR18, Koc16] constructions. That is normal in this area and not circular. The citation pattern is clean, and the examples in Section 3 look correct.\n\nRecommendation: send it to a serious referee, but require the referee to check Lemma 6.1 carefully. If the empty-interior argument is fixed, the paper should be accepted. I would not cite the main theorem until that happens.","headline":"Main theorem is the right result and the examples are sharp, but the empty-interior proof of Lemma 6.1 has a load-bearing gap that needs fixing.","tokens_in":33155,"tokens_out":7013,"would_cite":false,"duration_ms":61821,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["37E30","37E45","37B05"],"pacs":[],"model":"deepseek-v4-flash","headline":"For non-annular torus maps with small wandering domains, bounded rational drift is exactly the circle-factor condition.","keywords":["irrational circle factor","periodic point free","torus homeomorphism","rotational deviations","wandering domains","rho-centralized skew-product","pseudo-rotation","Kronecker factor"],"falsifier":"Search for a periodic-point-free, orientation-preserving, non-eventually-annular homeomorphism of $\\mathbb{T}^2$ that has small wandering domains and satisfies the bounded-deviation estimate (2) for some rational-slope vector $v$ but admits no irrational circle factor; a single such example would refute Theorem A. The paper's Section 3 examples show that removing the small-wandering-domains hypothesis already produces maps with bounded deviations and no factor, so the check must keep that hypothesis intact.","tokens_in":32194,"feed_emoji":"🌀","tokens_out":8294,"duration_ms":71479,"temperature":0.7,"pith_summary":"This paper proves a complete characterization of when a periodic-point-free homeomorphism of the 2-torus factors onto an irrational circle rotation. A necessary condition from earlier work is that the lift has uniformly bounded projections onto some rational-slope direction: the displacement along that direction differs from $n\\rho$ by a uniform constant. The paper shows that, for orientation-preserving non-eventually-annular maps with small wandering domains, this condition is also sufficient. Annularity and the geometry of the wandering set are therefore the only obstructions. The result extends earlier conservative results to general periodic-point-free homeomorphisms and, in particular, to maps isotopic to nontrivial Dehn twists.","feed_headline":"Bounded rotational drift unlocks irrational circle factors","feed_subtitle":"With small wandering domains and no annular obstruction, bounded drift in one rational direction is both necessary and sufficient.","key_machinery":"The central object is the $\\rho$-centralized skew-product $F:\\mathbb{T}\\times A\\to\\mathbb{T}\\times A$, adapted from earlier constructions to the Dehn-twist isotopy class. It separates the rotational motion of $f$ from its sublinear deviations: boundedness of an $F$-orbit in the annulus coordinate is equivalent to bounded rotational deviations of the original orbit, and under $\\Omega$-recurrence the non-wandering points of $F$ correspond exactly to those of $f$. Theorem 5.6 shows that a bounded connected $F$-invariant open set meeting $\\Omega(F)$ has fibers whose complements in the annulus have exactly two unbounded connected components; the boundaries of these fibers are the essential annular continua used to build the semi-conjugacy.","core_discovery":"Theorem A states that an orientation-preserving, non-eventually-annular homeomorphism $f$ of $\\mathbb{T}^2$ having small wandering domains admits an irrational circle rotation as a topological factor if and only if there is a rational-slope vector $v\\neq 0$ such that, for any lift $\\tilde{f}$, there exist $\\rho\\in\\mathbb{R}\\setminus\\mathbb{Q}$ and $C>0$ with $|\\langle \\tilde{f}^n(z)-z, v\\rangle - n\\rho|\\leq C$ for all $z\\in\\mathbb{R}^2$ and $n\\in\\mathbb{Z}$. The necessity is classical; the sufficiency is proved by constructing, from bounded connected invariant open sets of the $\\rho$-centralized skew-product, a family of pairwise disjoint essential annular continua $C^s\\subset A$ indexed by $s\\in\\mathbb{R}$ that satisfy $\\hat{T}_1(C^s)=C^{s+1}$ and $\\hat{f}(C^s)=C^{s+\\rho}$. These continua are the fibers of the semi-conjugacy; their equivariance under the vertical translation and under $\\hat{f}$ yields a continuous map $h:\\mathbb{T}^2\\to\\mathbb{T}$ with $h\\circ f = T_\\rho\\circ h$.","pith_inferences":["A natural testable extension is whether $\\Omega$-recurrence alone, without the diameter control of small wandering domains, already forces the factor; the diameter estimates enter only in proving that the constructed invariant sets stay bounded and that their translates avoid $\\Omega(F)$.","The semi-conjugacy produced by the proof gives an invariant one-dimensional pseudo-foliation of $\\mathbb{T}^2$ whose leaves have a well-defined homological direction, so the result supplies a tool for questions about invariant foliations and the structure of periodic-point-free torus maps with wandering points.","If the bounded-deviation condition is to be checked in practice, it can be computed along a single rational direction by sampling lifts of long orbits; persistent boundedness of $|\\langle \\tilde{f}^n(z)-z,v\\rangle - n\\rho|$ would then be strong numerical evidence for an irrational circle factor, provided the wandering domains are small."],"forward_implications":["Corollary 1.1: every totally irrational pseudo-rotation with uniformly bounded rotational deviations and small wandering domains is a topological extension of the corresponding minimal translation of $\\mathbb{T}^2$.","Corollary 1.2: a periodic-point-free homeomorphism isotopic to a nontrivial $k$-Dehn twist with small wandering domains admits an irrational circle factor exactly when its vertical deviations are uniformly bounded.","Combined with the classification of minimal Kronecker factors of torus homeomorphisms, the result gives a full description of which periodic-point-free homeomorphisms of $\\mathbb{T}^2$ have nontrivial Kronecker factors: any such factor is an irrational circle rotation or a totally irrational torus rotation assembled from two transversal circle factors.","The three examples in Section 3 demonstrate that the small-wandering-domains hypothesis is sharp: with lift-unbounded, fully essential, or uniformly large wandering domains, uniformly bounded rotational deviations can hold while no irrational circle factor exists."],"supporting_citations":[{"why":"Supplies the classical necessity lemma linking irrational circle factors to uniformly bounded rational-direction deviations, and the area-preserving version of the characterization that this paper extends.","marker":"[JT17]"},{"why":"The principal prior theorem that is extended, and the source of the annular-continuum construction used to build the semi-conjugacy.","marker":"[Jäg09]"},{"why":"Introduces the $\\rho$-centralized skew-product and its invariant pseudo-foliations, here adapted to Dehn-twist isotopy classes.","marker":"[KPR18]"},{"why":"Provides the orbit-of-open-set and essential-point framework used in the recurrence and deviation lemmas.","marker":"[KT14]"},{"why":"Classifies minimal Kronecker factors of torus homeomorphisms, reducing the search for nontrivial Kronecker factors to circle rotations and totally irrational torus rotations.","marker":"[HJ17]"},{"why":"Justifies the assumption, used in the final construction, that the semi-conjugacy can be chosen with annular-continuum fibers.","marker":"[JP15]"}],"fun_headline_variants":["No periodic points, bounded drift: irrational factor appears","Bounded drift in rational direction forces irrational circle factor","Annularity and wandering domains: only obstructions to irrational factors","Rational drift bound implies irrational circle factor on torus"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that every connected component of the wandering set is lift-bounded and only finitely many have diameter above any fixed threshold, since this is what forces $\\Omega$-recurrence and keeps the invariant constructions bounded; if a wandering domain is lift-unbounded or an infinite family has large diameter, the conclusion can fail, as the paper's examples show.","fun_headline_variants_meta":{"raw":{"variants":["No periodic points, bounded drift: irrational factor appears","Bounded drift in rational direction forces irrational circle factor","Annularity and wandering domains: only obstructions to irrational factors","Rational drift bound implies irrational circle factor on torus"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000336,"raw_usage":{"total_tokens":1842,"prompt_tokens":906,"completion_tokens":936,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":522,"completion_tokens_details":{"reasoning_tokens":870}},"tokens_in":522,"tokens_out":936,"duration_ms":7760,"temperature":1.0,"reasoning_tokens":870,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:05:05.753722+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Search for a periodic-point-free, orientation-preserving, non-eventually-annular homeomorphism of $\\mathbb{T}^2$ that has small wandering domains and satisfies the bounded-deviation estimate (2) for some rational-slope vector $v$ but admits no irrational circle factor; a single such example would refute Theorem A. The paper's Section 3 examples show that removing the small-wandering-domains hypothesis already produces maps with bounded deviations and no factor, so the check must keep that hypothesis intact.","supporting_citations":[],"review_version":1}