{"id":"8f3bdd92-0426-4fd9-99aa-fd9a3a67d695","arxiv_id":"1908.06145","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Cycles with coprime over-rotation pair are over-twists exactly when the forced over-rotation interval starts at their over-rotation number; for non-coprime pairs, very badly ordered cycles exist without block structure over over-twists.","lead":"This mathematics paper studies cycles of interval maps and their over-rotation numbers, which measure how often an orbit jumps across a fixed point. It proves a characterization for coprime cycles and constructs surprising very badly ordered cycles for non-coprime cases, showing interval maps differ from circle maps.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Section 5 proves r=rho for g|P via code non-decreasing, but never checks the second clause of Definition 3.2: no block structure over an over-twist of the same over-rotation number. Without that check the central existence claim is unsupported.","rationale":"The reader's weakest assumption identifies the same missing verification: Section 5 establishes that g|P has the correct over-rotation interval but not that it has no block structure over an over-twist of the same over-rotation number. I agree that this is the most load-bearing concern. The paper's own text supports the gap: Definition 3.2 has two clauses; the final paragraph of Section 5 says that the code computation 'verifies our claim in two different ways,' but both ways only show Ig=[p/q,1/2] (equivalently r=rho). Neither way addresses block structure. The concern is not a disagreement with the mathematical community's consensus; it is an internal proof gap. It is also not merely cosmetic, because the construction is formed by gluing k over-twist cycles of period q, and a block structure over gamma_{p/q} is a concrete alternative outcome that would satisfy every verified property while making the constructed orbit merely badly ordered, not very badly ordered. The proposed test -- running the block-structure reduction on the explicit permutation from Example 5.1 -- would settle the question directly. Since the reader's CONDITIONAL verdict already asks for exactly this missing verification, no verdict adjustment is needed; the concern confirms rather than changes the conditional assessment.","tokens_in":16593,"tokens_out":17088,"duration_ms":157596,"concrete_test":"Implement the Section 5 construction literally for a small case, say (k,p,q)=(2,2,5), from the stated gluing rules: g(c13)=c21, g(c2(l-1))=c31, etc., with the spatial order given by the nested lifting. Compute the resulting kq-point permutation of {1,...,kq}. Then run the standard block-structure reduction of [MN90] on this permutation. If the reduced irreducible pattern has period q=5 and over-rotation pair (2,5), and its code is strictly monotone, then g|P has a block structure over the over-twist gamma_{2/5}, contradicting Definition 3.2 and invalidating the Section 5 claim. If the reduction is trivial (the full 10-cycle has no invariant partition into consecutive blocks), then the missing no-block-structure condition is confirmed for this case; repeat for another parameter triple to test the general claim.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's headline construction in Section 5 claims that g|P is the desired unimodal very badly ordered periodic orbit, but it verifies only part of Definition 3.2. The final verification computes that the kneading sequence of g|P lies between nu'_{p/q} and nu_{p/q}, and/or that the code is non-decreasing, and concludes Ig=[p/q,1/2]. This establishes r_{g|P}=rho(g|P), but Definition 3.2 also requires that g|P has no block structure over an over-twist of over-rotation number p/q. That second condition is never checked. The gap is load-bearing because a non-decreasing code with equalities is exactly what a block structure over an over-twist produces: chasing through the definitions, if pi has a block structure over a pattern zeta, then r_pi=r_zeta and rho(pi)=rho(zeta), and if zeta is an over-twist, the blocks collapse to points with monotone code while the full code is non-decreasing with equalities within blocks. Lemma 2.2 even shows that equal code values force the over-rotation pair to be non-coprime, which is compatible with both possibilities. Moreover, the construction starts with k copies of the over-twist gamma_{p/q} and merges them, so a block structure over gamma_{p/q} is a natural candidate; the paper gives no argument excluding it. If g|P were such a blow-up, all the properties actually verified in Section 5 would hold, but the claimed new phenomenon of very badly ordered interval patterns would not exist for the constructed orbits.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies over-rotation numbers of interval maps. It proves that for a convergent periodic orbit pattern with coprime over-rotation pair, equality between the left endpoint of the forced over-rotation interval and the over-rotation number forces the pattern to be an over-twist (Theorem 2.4). It then introduces 'very badly ordered' patterns, develops a criterion via non-decreasing code, constructs the strongest unimodal pattern associated with a rational over-rotation interval in Section 4, and in Section 5 claims to construct, for every non-coprime over-rotation pair (kp,kq) with 2p<q, a unimodal very badly ordered pattern by merging k copies of the over-twist gamma_{p/q}. The announced conclusion is that interval dynamics differ from the circle-map case: a non-coprime pattern can have r_pi = rho(pi) without a block structure over an over-twist.","tokens_in":16925,"tokens_out":12473,"duration_ms":115351,"significance":"Theorem 2.4 and the code criterion are clean and potentially useful, and the proof of Theorem 2.4 is transparent. The explicit description of the strongest kneading sequence nu'_rho and the algorithmic construction in Section 5 are concrete and would provide a family of examples not previously in the literature. The paper builds on established results and offers a falsifiable construction rather than a parameter fit. However, the central existence claim for very badly ordered patterns is not fully verified as written: Section 5 proves the equality of the left endpoint of the forced interval but not the absence of block structure required by Definition 3.2. If that gap can be filled, the paper would establish a genuine and interesting contrast with the circle-map theory.","major_comments":[{"comment":"The verification of the main claim checks only part of Definition 3.2. It shows that g|P is unimodal, has over-rotation pair (kp,kq), satisfies nu'_{p/q} ≻ nu_g ≻ nu_{p/q}, and has non-decreasing code, from which the authors infer I_g = [p/q,1/2]. It never checks the second defining condition of 'very badly ordered': that g|P has no block structure over an over-twist of over-rotation number p/q. This condition is not a consequence of I_g = [p/q,1/2] or of non-decreasing code; indeed, a pattern with a block structure over the over-twist gamma_{p/q} would also have those properties. Since the construction begins with k disjoint copies of gamma_{p/q} and then glues them into one orbit, a block structure over gamma_{p/q} is a natural candidate, and no argument in the paper rules it out. The proof of the main existence theorem is therefore incomplete.","section":"Section 5, paragraphs after the construction of g|P"},{"comment":"The example asserts, without computation or proof, that the displayed pattern pi 'does not have a block structure over an over-twist periodic orbit'. The only evidence given is a statement about the code being non-decreasing, which by Corollary 3.4 is equivalent to r_pi = rho(pi) but does not preclude a block structure over an over-twist. Since this is the first proposed example of a very badly ordered pattern, the missing verification is a substantive gap.","section":"Example 3.5"},{"comment":"The identification of nu'_rho as the kneading sequence of the constructed map is asserted through phrases such as 'straightforward verification', 'it easily follows from the dynamics', and 'because of periodicity'. The proof of Theorem 4.1 and the comparisons in Section 5 depend on this identification. In particular, the claim that the point c3 under g has itinerary equal to nu'_rho and the assertion that any orbit of over-rotation number less than p/q would force a kneading sequence stronger than nu'_rho are not demonstrated. A rigorous proof should be supplied.","section":"Section 4.2, construction of gamma'_rho and proof of Theorem 4.1"}],"minor_comments":[{"comment":"The notation c13 and c2(l-1) is ambiguous; it should be c_{1,3} or explained explicitly, otherwise the reader cannot distinguish c13 from c_{13} in the temporal labelling.","section":"Section 5, notation"},{"comment":"The sentence 'Hence by Theorem 1.2 I_g ⊃ [p/q, 1/2]' refers to a 'Theorem 1.2' that does not exist; the intended reference is likely Lemma 1.2 or a result from [BM97].","section":"Section 5, reference to Lemma 1.2"},{"comment":"The listed code values are not sufficient for the reader to check non-decreasingness without reconstructing the full spatial order and the map g|P; a table or explicit permutation for the example would help.","section":"Example 5.1, verification of code non-decreasing"},{"comment":"There are several typographical errors in the bibliography: 'Corrollary' in the text, 'mappimg' in [Sha64], 'aplications' in [BS13], and the citation label [MN90] does not match the listed authors Bobok and Kuchta.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The main concern is the missing proof of the no-block-structure condition in Section 5. This is not a presentational issue: without that check the central existence claim is unsupported. If the authors can prove that the constructed orbits have no block structure over an over-twist, the paper could become acceptable after the other technical points are addressed. If the construction in fact has such a block structure, the main theorem would be false and a different construction would be needed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe one thing to know: the coprime half of the paper is real and worth your time; the non-coprime half, which is the headline counterexample to the circle-map analogy, is not yet supported. Theorem 2.4 — if the over-rotation pair is coprime and the P-linear map has left endpoint equal to the over-rotation number, then the pattern is an over-twist — is a clean consequence of the code lemmas, and the proof is convincing. Corollary 3.4's code-non-decreasing criterion is also a useful tool.\n\nWhat's new: the very badly ordered definitions and the claim that they exist for every non-coprime pair (kp,kq) with 2p<q. The construction in Section 5, merging k copies of gamma_{p/q} into one unimodal orbit, is clever, and the explicit (4,10) example gives a concrete picture. Section 4's description of the strongest kneading sequence nu'_rho for a given over-rotation interval looks new and is plausibly correct.\n\nThe soft spot is exactly where the stress-test note lands. Definition 3.2 has two clauses: rho(pi)=r_pi and no block structure over an over-twist of the same number. Section 5 verifies the first: it shows the kneading sequence lies between nu' and nu, or equivalently that the code is non-decreasing, hence I_g=[p/q,1/2]. It never checks the second. The concern is not pedantry: a non-decreasing code with equalities is precisely what a block structure over an over-twist would produce, and the construction begins with k copies of gamma_{p/q}, so a block structure is a natural possibility. The equal code values visible in Example 5.1 make this live. Unless the authors prove irreducibility, the existence claim for very badly ordered patterns is unsupported. Example 3.5 has the same problem in miniature: it asserts the no-block-structure property without computation.\n\nIs the gap fatal? Not necessarily. The merged orbit may well be irreducible; the authors just haven't shown it. As written, the paper's main contrast with the circle-map results of [ALM98] rests on an unproven structural property. The coprime theorem and the kneading-sequence work are enough to justify peer review, but the referee should insist on a real proof of the no-block-structure clause before publication.\n\nI'd bring it to the reading group: good theorem, instructive gap, and a useful test of how much structural verification a construction paper owes. I'd cite the coprime part now; I'd wait on the very badly ordered existence until the gap is closed.\n\nRecommendation: send to peer review, expect heavy revision.","headline":"The coprime characterization is solid and new; the non-coprime existence result for very badly ordered patterns is missing the key no-block-structure verification.","tokens_in":17429,"tokens_out":20265,"would_cite":true,"duration_ms":178529,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["37E05","37E15","37E45"],"pacs":[],"model":"deepseek-v4-flash","headline":"A convergent pattern with coprime over-rotation pair is an over-twist exactly when its forced over-rotation interval starts at its own over-rotation number; non-coprime pairs break this equivalence and yield 'very badly ordered' patterns.","keywords":["over-rotation number","over-rotation pair","over-twist pattern","very badly ordered","unimodal map","kneading sequence","block structure","interval dynamics"],"falsifier":"For the constructed orbit with over-rotation pair $(kp,kq)$, compute whether its points can be partitioned into blocks that collapse to an over-twist of over-rotation number $p/q$; exhibiting such a block structure would disprove the 'very badly ordered' claim. A more direct check is to compute $r_\\pi$ by iterating the P-linear map and testing whether it is strictly less than $\\rho(\\pi)$, which would also falsify the construction.","tokens_in":16390,"feed_emoji":"🔄","tokens_out":5398,"duration_ms":47630,"temperature":0.7,"pith_summary":"The paper proves a sharp characterization for coprime over-rotation pairs: a convergent pattern whose P-linear map has over-rotation interval $[\\rho(P),1/2]$ must be an over-twist, and conversely every over-twist has this property. The new direction is the 'if' part. Once the over-rotation pair is not coprime, the paper shows the characterization fails: it constructs unimodal patterns with over-rotation number equal to the left endpoint of the forced over-rotation interval but with no block structure over an over-twist of the same rotation number. These 'very badly ordered' examples matter because they separate interval-map over-rotation behavior from classical circle-map rotation-number behavior, where the analogous equality would force a block structure over a rotation.","feed_headline":"Coprime cycles are over-twists iff they force no smaller rotation","feed_subtitle":"For non-coprime pairs the rule collapses, producing 'very badly ordered' interval patterns with no circle-map analogue.","key_machinery":"The code of a periodic orbit is a real-valued function $L$ defined by $L(f(y))=L(y)+\\rho-\\phi_a(y)$, where $\\rho$ is the over-rotation number and $\\phi_a$ records which side of the unique fixed point $a$ a point lies on; monotonicity of this code characterizes over-twists. The paper also uses the P-linear map, whose cycles are exactly the patterns forced by the original pattern, and kneading sequences of unimodal maps, through which it identifies the strongest unimodal pattern with a given over-rotation interval. The construction of very badly ordered patterns merges $k$ copies of the unimodal over-twist $\\gamma_{p/q}$ into one orbit, then compares the resulting kneading sequence with the strongest kneading sequence for the interval $[p/q,1/2]$.","core_discovery":"The central claim is Theorem 2.4: for a convergent pattern with coprime over-rotation pair $(p,q)$, having $r_\\pi = \\rho(\\pi)$ — equivalently, having the P-linear map's over-rotation interval equal to $[p/q,1/2]$ — forces the pattern to be an over-twist. The proof runs through the code function: with a coprime pair, the code cannot have equal values on distinct points on the same side of the fixed point, and if the code ever decreased, Lemma 2.1 would produce a forced periodic orbit of smaller over-rotation number, contradicting $r_\\pi=\\rho(\\pi)$. Hence the code is strictly monotone, and by the known over-twist criterion the pattern is an over-twist. The paper then shows this theorem is sharp: for every non-coprime pair $(kp,kq)$ with $2p<q$, Section 5 constructs a unimodal pattern whose over-rotation number equals the left endpoint of its forced over-rotation interval, yet which has no block structure over an over-twist of the same over-rotation number — a 'very badly ordered' pattern.","pith_inferences":["The existence of very badly ordered patterns suggests that the forcing relation among non-coprime patterns is not determined by the over-rotation interval alone, even though the interval is the same as for an over-twist pattern.","A testable numerical signature of the constructed patterns is a kneading sequence that matches the strongest kneading sequence for $[p/q,1/2]$ up to position $kq-1$ and then shows a $C$ where the strongest sequence has an $L$; this could be checked for larger $k$.","One might expect similar 'very badly ordered' phenomena in multimodal interval maps, since the construction relies only on unimodal kneading theory, although the paper does not address that case."],"forward_implications":["For a pattern with coprime over-rotation pair, the equality $r_\\pi=\\rho(\\pi)$ completely characterizes over-twist patterns.","Any pattern with $r_\\pi=\\rho(\\pi)$ that is not an over-twist must have a non-coprime over-rotation pair.","Very badly ordered patterns show that over-rotation forcing on the interval is not fully analogous to rotation-number forcing for circle maps of degree one.","For every non-coprime over-rotation pair $(kp,kq)$ with $2p<q$, there exists a unimodal very badly ordered pattern with that pair.","The strongest unimodal pattern forcing the over-rotation interval $[p/q,1/2]$ has an explicit kneading sequence, obtained from the over-twist kneading sequence by replacing each occurrence of the fragment $CRL$ with $LRR$."],"supporting_citations":[{"why":"Introduces over-rotation pairs and over-rotation numbers, the forcing order among them, and the over-rotation interval; provides the foundational setting for the paper.","marker":"[BM97]"},{"why":"Supplies the criterion that a pattern is an over-twist if and only if it is convergent and has monotone code, used in the proof of Theorem 2.4.","marker":"[BM99]"},{"why":"Proves that the only unimodal over-twist pattern of a given over-rotation number is $\\gamma_\\rho$, which underlies the kneading-sequence comparisons in Sections 4 and 5.","marker":"[BS13]"},{"why":"Defines badly ordered cycles for circle maps, the analogue that motivates the interval notion and provides the contrast the paper highlights.","marker":"[ALM98]"},{"why":"Provides the chain and loop lemmas for interval maps and the theory of P-linear maps and forcing, used throughout the proofs.","marker":"[ALM00]"},{"why":"The source of the code construction for periodic orbits of interval maps.","marker":"[BK98]"},{"why":"Gives the definition and basic properties of block structure and unique irreducible patterns, needed for the notion 'no block structure over an over-twist'.","marker":"[MN90]"}],"fun_headline_variants":["Non-coprime cycles break the over-twist rule","Very badly ordered cycles exist with no block structure","Over-rotation criterion fails for non-coprime pairs","No circle-map analogue: very badly ordered cycles","Non-coprime pairs yield very badly ordered patterns"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The Section 5 construction proves that the merged orbit has over-rotation interval $[p/q,1/2]$, but it never proves the second defining condition of 'very badly ordered': that the orbit has no block structure over an over-twist of the same over-rotation number.","fun_headline_variants_meta":{"raw":{"variants":["Non-coprime cycles break the over-twist rule","Very badly ordered cycles exist with no block structure","Over-rotation criterion fails for non-coprime pairs","No circle-map analogue: very badly ordered cycles","Non-coprime pairs yield very badly ordered patterns"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001093,"raw_usage":{"total_tokens":4566,"prompt_tokens":952,"completion_tokens":3614,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":568,"completion_tokens_details":{"reasoning_tokens":3538}},"tokens_in":568,"tokens_out":3614,"duration_ms":23804,"temperature":1.0,"reasoning_tokens":3538,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:55:15.530013+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"For the constructed orbit with over-rotation pair $(kp,kq)$, compute whether its points can be partitioned into blocks that collapse to an over-twist of over-rotation number $p/q$; exhibiting such a block structure would disprove the 'very badly ordered' claim. A more direct check is to compute $r_\\pi$ by iterating the P-linear map and testing whether it is strictly less than $\\rho(\\pi)$, which would also falsify the construction.","supporting_citations":[],"review_version":1}