{"id":"b6d252b9-4a11-496c-91bd-b10760f5f86f","arxiv_id":"2411.18360","paper_version":1,"verdict":"REJECT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"high","formal_verification":"none","parameter_count":0,"one_line_summary":"A distal annulus homeomorphism with invariant circle foliation is conjugate to a rotation whenever the foliation admits a continuous transversal, and non-linearizable examples exist for every irrational rotation number.","lead":"The paper asks whether distal, periodic-point-free homeomorphisms of an annulus with an invariant circle decomposition are always conjugate to rotations. It proves that a continuous transversal guarantees linearizability, and that for every irrational rotation number there exist non-linearizable examples built by folding rotations near the boundary.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Final step of Theorem 1.4 is unsupported: a folding circle with MFA>π/3 in each A_n does not by itself exclude a continuous transversal, and each local restriction actually admits one.","rationale":"The reader's main concern is correct and is the central problem: the only support for non-linearizability in Theorem 1.4 is an unproved one-sentence claim that property (3) prevents a continuous transversal. My analysis strengthens this by observing that each block A_n is itself a homeomorphic copy of the radial-circle decomposition, so a local transversal always exists; the alleged global obstruction must involve the infinite nesting and the limit at C_2, which the paper never addresses. I disagree with the reader on one secondary detail: the MFA definition is not strictly inapplicable to simple closed curves, since the endpoint coincidence γ(0)=γ(1) satisfies the existential condition, but this only makes MFA a basepoint-dependent angular range rather than a genuine folding invariant, so the geometric intuition is still not supplied. No formal verification, reproducible code, or independent computation supports the key implication. Thus the reader's REJECT verdict is appropriate, and my stress test does not change it. If a correct proof of the no-transversal assertion is supplied, Theorem 1.4 could become viable, but as written the central existence claim is unsupported.","tokens_in":5441,"tokens_out":14104,"duration_ms":136989,"concrete_test":"Restrict to one block, say q=7. Let Φ_n = φ_n^{-1}∘H_n, a homeomorphism from the standard annulus onto A_n. Plot the curve Φ_n({θ=0, 1≤|z|≤2}) and check that it intersects each invariant circle φ_n^{-1}(H_n(C_r)) exactly once. If it does, the assertion that a folding circle with MFA>π/3 in each A_n rules out a continuous transversal is false locally, and the burden shifts to showing that no global choice of such local transversals can be made consistently across all n with a limit at C_2.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Theorem 1.4 depends on the last sentence of §3.3: 'But this follows from property (3) of H that there is a folding circle in each A_n whose maximal folding angle is greater than π/3.' No proof is given that one invariant essential circle with large angular variation prevents a continuous global section of the decomposition P. This is not a routine consequence. On an individual block A_n, the restriction g|A_n is conjugate, via the homeomorphism Φ_n = φ_n^{-1}∘H_n, to the rigid rotation R_α. Hence the restriction of P to A_n is the image of the standard radial-circle decomposition under Φ_n, and Φ_n applied to a radial segment is a continuous transversal for P|A_n. Therefore the obstruction, if real, must come from the way local transversals fail to glue across the nested annuli A_n and to converge at C_2; the paper contains no argument for this. A further clarification: the MFA definition in §3.1 is not literally inapplicable to simple closed curves, because γ(0)=γ(1) satisfies the displayed condition using endpoints, but then MFA measures angular distance from the chosen basepoint rather than a genuine self-intersection fold, and Claim 1's lower bound depends on choosing that basepoint at π(w2). The missing gluing/convergence proof is the load-bearing gap.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies distal, periodic-point-free homeomorphisms of the annulus. Building on the authors' previous structure theorem (Theorem 1.1), it proves Theorem 1.3: if the invariant-circle decomposition P admits a continuous section, then the map is topologically conjugate to a rigid rotation. It then claims Theorem 1.4: for every irrational α∈(0,1) there exists such a distal homeomorphism g with rotation number α that cannot be linearized, using a construction of annuli A_n accumulating at the outer boundary and homeomorphisms H_n that nearly commute with rational rotations. The non-linearizability is argued by asserting that the presence of a 'folding circle' with large maximal folding angle in each A_n prevents the existence of a continuous transversal for P.","tokens_in":5725,"tokens_out":10488,"duration_ms":101661,"significance":"The positive result Theorem 1.3 is a clean and plausible sufficient condition for linearizability, and it would be a useful complement to the structure theorem in [2]. If Theorem 1.4 were established, it would answer Question 1.2 negatively and provide the first examples of non-linearizable distal annulus homeomorphisms for every irrational rotation number. However, the proof of Theorem 1.4 in this manuscript has several load-bearing gaps, including an ill-defined quantity and an unsupported final step, so the significance of the paper as it stands is conditional. The paper does not provide machine-checked proofs or reproducible code; its contribution is purely theoretical.","major_comments":[{"comment":"The definition of MFA(γ) is vacuous for the curves to which it is applied. For an essential simple closed curve γ, the condition γ(t1)=γ(t3) with 0≤t1<t2<t3<1 cannot be satisfied, because such a curve is injective on [0,1). Thus MFA(γ) is the maximum over the empty set. In Claim 1 the inequality MFA(Γ)>π/3 is justified by comparing θ(π(w2)) and θ(π(w3)), but those are distinct points on the simple closed curve Γ, not a self-intersection. Consequently property (3) in Section 3.2, and the later use of folding circles, is not established.","section":"Section 3.1"},{"comment":"The construction of g is not shown to be well-defined on A\\C_2. The annuli A_n are closed and adjacent annuli share boundary circles, for example A_1∩A_2=C_{7/4} in the notation of the paper. Items (a) and (b) prescribe g on each A_n separately, without proving that the local definitions agree on the common boundaries C_{2-1/2^{n+1}}. The statement 'It follows from the construction that g is continuous on A \\ C_2' is therefore unsupported; compatibility on all shared interfaces must be checked first.","section":"Section 3.3, items (a) and (b)"},{"comment":"The assertion that a folding circle in each A_n with maximal folding angle greater than π/3 prevents a continuous transversal is stated in one sentence and is not a routine consequence. On each individual A_n, g is conjugate to R_α, so the restriction of P to A_n admits continuous local transversals. Any obstruction must come from the way local transversals fail to glue across the nested annuli and converge at C_2, and the paper contains no argument for this. This is a load-bearing gap in the proof of Theorem 1.4.","section":"Section 3.3, last paragraph"},{"comment":"Claim 6 asserts that x_{n_i}^α converges in C_α if and only if (n_i α) converges in S^1, citing Claim 3 and the normalization on C_{1.5}. For a general C_α this requires that the convergence of iterates under g|C_α is equivalent to convergence of n_i α, which is not proved here; if it follows from [2], the authors should state the precise result they are invoking. As written, this step in the construction of the conjugacy Ψ is too terse.","section":"Section 2, Claim 6"}],"minor_comments":[{"comment":"The phrase 'F or each irrational number' contains a typo and should read 'For each irrational number'.","section":"Theorem 1.4 statement"},{"comment":"The phrase 'squeezing along the radical direction' should be 'squeezing along the radial direction'.","section":"Section 3.3, item (a)"},{"comment":"The sentence 'There there are a, b, c, d ∈ Z ...' contains a duplicated 'There' and the notation (L_a,L_b), [L_a,L_b] is confusing; please clarify.","section":"Section 2, Claim 3 proof"},{"comment":"The phrase 'there is no a transversal' should be 'there is no transversal'.","section":"Section 3.3, final paragraph"},{"comment":"The text refers to FIGURE 1 in the proof of Section 3.2, but the figure is not visible in the manuscript text; if the figure is essential, it should have a caption and be clearly legible.","section":"Figure 1"},{"comment":"The assumption that we may take C_{1.5}={|z|=1.5} and g|C_{1.5} as a rigid rotation is used without justification; a sentence explaining why this normalization is harmless would be helpful.","section":"Section 2, beginning"}],"recommendation":"reject","confidential_remarks":"The main new result Theorem 1.4 is not established: the maximal folding angle is undefined for the simple closed curves to which it is applied, the construction of g lacks compatibility on shared boundaries of the A_n, and the final no-transversal step is asserted without proof. These are not cosmetic issues and I do not see how a modest revision of the current construction could repair them. Theorem 1.3 might be publishable as a separate contribution if Claim 6 is fully justified, but as submitted the paper should be rejected."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here is my take. Theorem 1.3 is a reasonable sufficient condition, and the proof mostly holds together: if the decomposition admits a continuous section, the conjugacy construction is standard, and the continuity arguments, while sketched, are plausible. The genuinely new thing is the attempt at Theorem 1.4: explicit folding homeomorphisms H_n satisfying (1)–(3), then a nested-annulus assembly. The construction is concrete and the estimates are checkable; that part is good.\n\nThe final step of Theorem 1.4, however, is a one-sentence assertion: the existence of a folding circle in each A_n with maximal folding angle greater than π/3 is said to rule out a continuous transversal. No proof is given. This is load-bearing, because the restriction of g to each A_n is conjugate to the rigid rotation R_α, so each block individually admits a continuous local transversal. The obstruction, if real, must come from how local transversals fail to glue across the nested annuli and converge at C_2; the paper contains no argument for that. I also think the MFA definition is not doing what the authors want: as written, for a simple closed curve it measures angular distance from the chosen basepoint via the endpoint coincidence, not a genuine fold, and Claim 1's lower bound depends on that basepoint choice. That is a weakness in the intuitive content, though not the main gap.\n\nIf the missing gluing argument can be supplied, the result would be a nice answer to Question 1.2. As it stands, the paper does not establish non-linearizability.\n\nWho is this for? People working on distal surface homeomorphisms and structure theorems for circle decompositions. The paper is short and the construction is transparent. I would send it to a referee, because the gap might be fixable and the question is worth resolving, but I would not accept it in its current form.","headline":"A nice sufficient condition and an explicit construction, but the non-linearizability proof stops one sentence short of the actual argument.","tokens_in":6212,"tokens_out":2468,"would_cite":false,"duration_ms":21695,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["37E30","37B05","37E45"],"pacs":[],"model":"deepseek-v4-flash","headline":"Periodic-point-free distal annulus homeomorphisms need not be conjugate to rigid rotations.","keywords":["distal homeomorphism","annulus","linearization","rotation number","continuous decomposition","transversal","folding rotation","periodic point free"],"falsifier":"Find a continuous arc in one of the constructed maps $g$ that meets every member of its invariant-circle decomposition exactly once; together with Theorem 1.3 this would give an explicit conjugacy to a rigid rotation and refute Theorem 1.4 for that $\\alpha$. A numerical or geometric search for such an arc across the nested annuli $A_n$, or a proof that every candidate arc is cut by one of the folds, would settle the disputed implication.","tokens_in":5204,"feed_emoji":"🔄","tokens_out":9701,"duration_ms":88716,"temperature":0.7,"pith_summary":"Distal means any two distinct points stay a positive distance apart under iteration; periodic-point-free means no orbit eventually repeats. For boundary-preserving homeomorphisms of the annulus with these properties, earlier work produced a continuous decomposition into invariant circles on each of which the map rotates by the same irrational number. The natural question is whether the whole map must be a rigid rotation in disguise. This note answers with a dichotomy: if that decomposition admits a continuous transversal, then the map is topologically conjugate to a rigid rotation; but for every irrational $\\alpha\\in(0,1)$, there is such a map with rotation number $\\alpha$ that is not conjugate to any rigid rotation. In other words, the rotation number alone does not force linearizability; the way the invariant circles are folded into the annulus matters.","feed_headline":"Folded annuli defeat rotation conjugacy at every irrational angle","feed_subtitle":"A transversal forces a distal annulus map to be a rotation; new examples show it can fail for every irrational alpha.","key_machinery":"For the positive result, the central object is the transversal $\\gamma$ of the decomposition $\\mathcal P$. The paper forms the leaves $L_n=g^n(\\gamma)$; distality makes them pairwise disjoint transversals, and the regions between consecutive leaves are decomposed continuously into arcs $C_{m,n}^\\alpha$ along invariant circles. This yields coordinates on the annulus and a map $\\Psi$ defined on $L_n$ by $x\\mapsto e^{2\\pi i n\\theta}\\psi(g^{-n}x)$, and continuity of the section is what makes $\\Psi$ a homeomorphism conjugating $g$ to $R_\\theta$. For the negative result, the central object is a periodic folding rotation $H$: a homeomorphism of the annulus that commutes with a rational rotation $R_{p/q}$, has controlled angular distortion (at most a factor $5q$ on angular differences along each circle), and contains a folding circle whose maximal folding angle, the largest angular gap between two preimages landing on the same folded point, exceeds $\\pi/3$. Squeezing such $H$ into thin annuli near the boundary and iterating with $R_\\alpha$ produces the non-linearizable examples.","core_discovery":"The paper's central claim is Theorems 1.3 and 1.4 together. Theorem 1.3 says that if the invariant-circle decomposition $\\mathcal P$ admits a continuous section $\\varphi:[0,1]\\to\\mathbb{A}$ meeting each circle exactly once, then $g$ is topologically conjugate to the rigid rotation $R_\\theta$ whose angle is the common irrational rotation number of the circle restrictions. The conjugacy is built from the iterates $L_n=g^n(\\varphi([0,1]))$ of the transversal. Theorem 1.4 says this hypothesis is not automatic: for every irrational $\\alpha\\in(0,1)$ there is a distal, boundary-preserving, periodic-point-free homeomorphism with rotation number $\\alpha$ that is not linearizable. These examples are made by placing, inside nested annuli $A_n$ accumulating at the outer boundary, conjugates of a folding homeomorphism $H_n$ that commutes with a rational rotation $R_{p_n/q_n}$, where $p_n/q_n\\to\\alpha$, while the rest of the annulus is a pure rotation. Each block contains a folded circle whose maximal folding angle exceeds $\\pi/3$, which the authors use to rule out any continuous transversal.","pith_inferences":["The paper leaves smoothness open; a natural next step is to try smoothing the folding blocks, since the bound $5q|\\alpha-p/q|$ suggests that Diophantine properties of $\\alpha$ will control whether folds can survive a $C^1$ or $C^\\infty$ perturbation.","The threshold $\\pi/3$ for the maximal folding angle comes from the specific geometry of the construction; the true minimal folding angle that destroys all continuous transversals could be smaller and is worth isolating.","The Theorem 1.3 proof uses only elementary properties of the continuous section, so the same 'section implies linearization' mechanism is a plausible template for other spaces carrying continuous decompositions into invariant circles, such as closed annuli or tori.","The construction suggests that, in the $C^0$ category, non-linearizable distal annulus homeomorphisms with a fixed irrational rotation number may be typical rather than exceptional, since folds can be inserted in arbitrarily thin annuli without changing the rotation elsewhere."],"forward_implications":["A continuous transversal for the invariant-circle decomposition is a sufficient condition for linearization, so any non-linearizable example must have a decomposition with no continuous section.","Non-linearizability is compatible with every irrational rotation number in $(0,1)$; well-approximable and poorly approximable $\\alpha$ behave the same at this level of regularity.","The obstruction is local in the radial direction: outside a sequence of thin annuli the constructed map is a rigid rotation, so global conjugacy can be destroyed by folding only near the boundary.","For any constructed example, exhibiting a continuous transversal would, by Theorem 1.3, give an explicit conjugacy to a rotation; excluding such transversals is therefore the decisive test of non-linearizability."],"supporting_citations":[{"why":"Supplies the structure theorem decomposing the annulus into invariant circles with a common irrational rotation number; Theorem 1.3 and the non-linearizability criterion are built on it.","marker":"[2]"},{"why":"Frames the smoothness question and its connection to Herman's question; the paper cites it when asking whether the non-linearizable examples can be made smooth.","marker":"[1]"}],"fun_headline_variants":["A transversal forces rotation, but folding beats it for every alpha","Distal annulus maps: section yields rotation, folded examples break it","For every irrational angle, a distal annulus map escapes linearization","No continuous section means no rotation conjugacy, for any irrational"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is the one-sentence assertion at the end of Section 3.3 that a folding circle with maximal folding angle greater than one-third of a full turn inside every annulus $A_n$ prevents any continuous transversal of the decomposition; if that implication is false, the constructed maps could still be linearizable.","fun_headline_variants_meta":{"raw":{"variants":["A transversal forces rotation, but folding beats it for every alpha","Distal annulus maps: section yields rotation, folded examples break it","For every irrational angle, a distal annulus map escapes linearization","No continuous section means no rotation conjugacy, for any irrational"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00016,"raw_usage":{"total_tokens":1254,"prompt_tokens":992,"completion_tokens":262,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":608,"completion_tokens_details":{"reasoning_tokens":188}},"tokens_in":608,"tokens_out":262,"duration_ms":3376,"temperature":1.0,"reasoning_tokens":188,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T11:18:18.512006+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find a continuous arc in one of the constructed maps $g$ that meets every member of its invariant-circle decomposition exactly once; together with Theorem 1.3 this would give an explicit conjugacy to a rigid rotation and refute Theorem 1.4 for that $\\alpha$. A numerical or geometric search for such an arc across the nested annuli $A_n$, or a proof that every candidate arc is cut by one of the folds, would settle the disputed implication.","supporting_citations":[{"cited_title":"The structure of periodic point free distal homeomorphisms on the annulus","cited_arxiv_id":"2406.10674","evidence_quote":"Supplies the structure theorem decomposing the annulus into invariant circles with a common irrational rotation number; Theorem 1.3 and the non-linearizability criterion are built on it."},{"cited_title":"Bramham, Z","cited_arxiv_id":null,"evidence_quote":"Frames the smoothness question and its connection to Herman's question; the paper cites it when asking whether the non-linearizable examples can be made smooth."}],"review_version":1}