{"id":"481c50f8-8c94-482f-ae1f-a26926a3bd70","arxiv_id":"2411.14312","paper_version":4,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For interval translation maps on any fixed number of intervals, the stable, finite-type maps form an open and dense subset of parameter space.","lead":"This paper proves that interval translation maps, which slide pieces of a line, can always be perturbed a tiny amount into stable maps whose long-term behavior is finitely many looping intervals. This settles the topological version of a 1995 conjecture about how rare more complicated maps are.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 4.6 is only proved in a generic case; the omitted boundary and non-landing cases, asserted as 'analogous', are load-bearing for the density theorem.","rationale":"The reader's weakest assumption correctly identifies the unrestricted validity of Theorem 4.6 as the most load-bearing point. The paper's proof of Theorem 4.6 explicitly covers only the case where boundary points of J0 land on discontinuities and each relevant point lands on at least two discontinuities, with the other cases dismissed as analogous or simpler. Since Corollary 4.8, the stability characterization (Theorem 5.12), and the approximation of eventually periodic maps (Theorem 6.1) all depend on the full statement, an unproved special case is a genuine gap, not a stylistic omission. The concern is internal to the proof rather than a disagreement with consensus: it is a missing argument, not a contradictory claim. The proposed test—writing out the omitted case or running exact-arithmetic independence checks—would settle whether the gap is benign. In good faith, the paper contains substantial independent support: the density of eventually periodic maps is proved cleanly via rational parameters, and the stability theory is coherent. The correct verdict remains CONDITIONAL, as the reader stated, asking for completion or verification of the omitted cases. I therefore do not change the reader's verdict.","tokens_in":59681,"tokens_out":3811,"duration_ms":34908,"concrete_test":"Re-derive the proof of Theorem 4.6 in the case where the left boundary point x^+ of J0 does not land on a discontinuity before its return to J0, following the partition-refinement arguments of Sections 4.2–4.4, and check whether conclusions (3) and (4) still force the vanishing needed for Corollary 4.8. In parallel, run an exact-arithmetic computer search over rational ITMs with r=3 and r=4: for each map, compute the component interval J0, the landing/return vectors defined in Definitions 4.1–4.3, and test the linear independence of the Corollary 4.8 vectors, while systematically including examples where x^+ misses all discontinuities and where some a_j lands on exactly one discontinuity. If any non-zero dependence is found, the density theorem is false; if all tested examples show independence, the 'analogous cases' claim gains strong computational support.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 4.1 (paragraph after Definition 4.3) restricts the proof of Theorem 4.6 to the case where the boundary points of the distinguished interval J0 land on discontinuities of T and where every relevant point a_j^{0,±} lands on at least two discontinuities before returning to J0. The remaining cases are explicitly declared 'analogous' or 'simpler' but no proof is provided. This is load-bearing because Theorem 4.6 is the sole input to Corollary 4.8, the linear independence of the vectors L, C, R, which is then used in Proposition 5.16, Theorem 5.22, Lemma 6.7, Theorem 6.1, and Lemma 7.8. If any of the omitted cases admitted a non-trivial linear dependence between the vectors of Corollary 4.8, the perturbation steps in Sections 5 and 6 could fail exactly where they need to remove a critical connection, so the density of stable maps would not follow. The paper gives no indication that the omitted cases have been checked independently, and the same pattern reappears in Section 8.1 when Theorem 4.6 is invoked for the extended parameter space with β_0^+ and β_r^- added as parameters.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves a topological version of the Boshernitzan–Kornfeld conjecture for interval translation maps: for every r ≥ 2, stable ITMs are dense in the parameter space ITM(r), and consequently finite type maps contain an open dense set while infinite type maps have empty interior. The proof is organized around three theorems: eventual periodic maps are dense (Theorem A); for finite type maps stability is equivalent to the absence of critical connections together with Matching (Theorem B); and every eventually periodic map can be perturbed to a stable map (Theorem C). The central technical input is Theorem 4.6, a linear-independence statement for dynamically defined landing, return, and critical-connection vectors, which is used throughout Sections 5, 6, and 7. The paper closes with a discussion of the Bruin–Troubetzkoy family and several open problems.","tokens_in":59906,"tokens_out":10062,"duration_ms":97182,"significance":"If correct, the paper resolves a natural topological strengthening of the Boshernitzan–Kornfeld conjecture and introduces a stability theory for ITMs that is likely to be influential. The proof is largely self-contained, and the main theorems are proved by direct perturbation constructions rather than by renormalization or by importing external results. The three-step architecture (density of eventually periodic maps, stability equivalence, approximation of eventually periodic maps) is clean and convincing in outline. The paper also explicitly identifies open questions about the geometry of stable regions, which should stimulate further work. However, the central linear-independence theorem is proved only under simplifying assumptions, and several later arguments rely on the full statement in ways that the printed text does not justify. The result is therefore plausible but not yet fully verified as written.","major_comments":[{"comment":"Theorem 4.6 is proved only in the case where the boundary points of the distinguished interval J0 land on discontinuities and where every point a_j^{0,±} lands on at least two discontinuities before returning to J0. The remaining cases are declared 'analogous' or 'simpler', but no proof is given. This is load-bearing because Corollary 4.8, which is the form in which Theorem 4.6 is used, is invoked in Proposition 5.16, Theorem 5.22, Lemma 6.7, Theorem 6.1, and Lemma 7.8. If any omitted case admitted a nontrivial dependence, the perturbation steps that remove critical connections could fail, and the density theorem would not follow. The revision should either supply complete proofs for the omitted cases or state precisely which version of Theorem 4.6 is proved and verify that this version is sufficient for every later application.","section":"Section 4.1, paragraph after Definition 4.3"},{"comment":"The perturbation argument in the A3 part of the proof is not fully specified. The text prescribes values ⟨v_j, δ⟩ = −ϵ for odd j and +ϵ for even j and invokes linear independence of the vectors, but if the same landing vector appears more than once along the ghost cycle, linear independence of the set does not permit assigning different dot products to identical vectors. Moreover, with the stated alternating signs, the total prescribed displacement around a cycle is not zero, so it is not clear that the orbit closes up to a periodic point as claimed. Please rewrite this step with explicit indexing of the ghost-tree path, a statement of which vectors are being solved for, and a verification that the resulting perturbation makes β_{i_1}^+ periodic in the sense required to enlarge X.","section":"Theorem 5.22, proof of A3"},{"comment":"The extension of Theorem 4.6 to the parameter space where β_0^+ and β_r^- are free parameters is asserted but not proved. These points are not discontinuities, so Definitions 4.1–4.4 do not directly associate landing, return, or critical-connection vectors to them. The statement 'The proof remains unchanged otherwise' is not a proof, and Theorem 8.1 depends on this extension. The revision should give the modified statement of Theorem 4.6 for the extended parameter space and indicate which parts of the proof of Section 4 carry over verbatim and which require change.","section":"Section 8.1, Theorem 8.1"},{"comment":"The proof of Lemma 7.8 again invokes Theorem 4.6 in a boundary-point case that is not covered by the printed proof: when both boundary points of the return interval do not land on discontinuities, the argument reduces the number of independent vectors from n−1 to n−2 and says it is 'simple to check' that the larger set is linearly dependent for n=3 and n=2. Since Lemma 7.8 is used in Lemma 7.7 and Theorem 7.1, this reduction should be written out explicitly, and the needed case of Theorem 4.6 should be included among the omitted cases supplied in the revision.","section":"Lemma 7.8"}],"minor_comments":[{"comment":"The definition of ghost preimage contains an internal variable confusion: it says a discontinuity β_*^- that lands on β_*^- is a ghost preimage of β_*^+, but the two occurrences of β_*^- should refer to different discontinuities, with the landing target being the −-side of the signed point corresponding to β_*^+. Please rewrite the clause so that the source and target of the landing are unambiguous.","section":"Definition 5.7"},{"comment":"The displayed identity involving C^{i,−}(1, m_i^{−1}) has mismatched indices: the final index is written as m_i^- in one place and m_i^+ in the summation range, and the expression is not syntactically well formed. Please correct the indexing so that the identity can be checked.","section":"Lemma 6.7"},{"comment":"In the rescaling step after the proof of Theorem 8.1, the text says that multiplying all parameters by 1/β_1^- gives that the left boundary point is now 1^+. This should presumably refer to normalizing the right endpoint β_r^- to 1, and the label β_1^- appears to be a typo for β_r^-.","section":"Section 8.1, rescaling argument"},{"comment":"The list of translation factors is written as γ_1, γ_2, ..., γ_3 in the first displayed paragraph of Section 1.1; it should be γ_1, ..., γ_r.","section":"Introduction, Section 1.1"},{"comment":"The text refers to 'Figure 8.1' but the figure illustrating the Bruin–Troubetzkoy triangles appears earlier; please recheck the figure numbering or add an explicit cross-reference.","section":"Section 8.2"}],"recommendation":"major_revision","confidential_remarks":"The central claim is likely correct and the paper is a serious contribution, but the printed proof of the key linear-independence theorem covers only a generic case, and the omitted cases are used in several later arguments. I would not reject the paper on this basis, but I would require the authors to supply the missing cases or to adjust the statements and later uses accordingly. The revision should also clarify the perturbation constructions in Theorem 5.22 and Lemma 7.8, where the published text is too terse for the claims being made."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Let me give you the short version: this is a serious paper that very likely proves the topological Boshernitzan–Kornfeld conjecture for all r, but the printed proof of the central transversality theorem (Theorem 4.6) only covers a generic case and declares the remaining cases analogous. That gap is load-bearing, and until it is filled the density theorem is conditional.\n\nWhat is genuinely new: the stability framework (ACC + Matching), the linear-independence of itinerary vectors, and the three-step strategy (eventually periodic maps are dense; stable ⇔ ACC+Matching; stable maps approximate eventually periodic maps). The paper gives a detailed, self-contained proof of the main steps, and I saw no sign of circularity or data-fitting. The structure is honest and the exposition is careful.\n\nThe soft spot is precisely where the reader's report points: Section 4.1 assumes the boundary points of J0 land on discontinuities and every a_j lands on at least two discontinuities. The other cases are dismissed as simpler or analogous. Corollary 4.8 then feeds into Proposition 5.16, Theorem 5.22, Lemma 6.7, and Theorem 6.1, so a counterexample in an omitted case would break the approximation argument. The same pattern reappears when Theorem 4.6 is invoked in the extended parameter space of Section 8.1.\n\nI should say the concern is not that I found an error—I didn't. It is that the proof as written is incomplete at a load-bearing point. A referee should ask for a complete treatment of the omitted cases, or a convincing reduction, before accepting. Given the size of the paper and the importance of the result, that is a reasonable request, not a nitpick.\n\nWho should read it: anyone working on piecewise isometries, ITMs, or the density-of-hyperbolicity analog for 1D systems. It deserves a serious refereeing process. I would bring it to a reading group once the gap is closed; right now I'd maybe mention it.\n\nRecommendation: send it to a good referee, but tell the referee to check Theorem 4.6 carefully.","headline":"A serious and likely correct proof of the topological Boshernitzan–Kornfeld conjecture for all r, but the load-bearing Theorem 4.6 is only proved in a generic case with the remaining cases relegated to 'analogous' arguments.","tokens_in":60421,"tokens_out":2050,"would_cite":false,"duration_ms":18737,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["37E05","37C20","37B20"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that for every number r≥2 of intervals, the stable interval translation maps—those of finite type whose first return maps are circle rotations—are dense in the space of all interval translation maps.","keywords":["interval translation maps","stable maps","finite type","infinite type","matching condition","absence of critical connections","transversality","piecewise isometries"],"falsifier":"Find an ITM in one of the cases Theorem 4.6 leaves out—say a component interval whose boundary points do not land on discontinuities—whose landing, return, and critical-connection vectors are linearly dependent with non-zero coefficients; the approximation argument would then fail at that point, since every later perturbation is built on that independence.","tokens_in":59488,"feed_emoji":"🔄","tokens_out":6504,"duration_ms":60716,"temperature":0.7,"pith_summary":"Interval translation maps (ITMs) are piecewise translations of an interval obtained by dropping the bijectivity assumption from interval exchange transformations. This paper proves a topological prevalence result: for every r≥2, stable ITMs form a dense subset of the full parameter space ITM(r), and because stable maps are automatically of finite type, finite-type maps contain an open dense set while infinite-type maps have empty interior. If true, this settles the topological version of a long-standing measure-theoretic conjecture from 1995. Stable here means the non-wandering set varies continuously in the Hausdorff topology and is homeomorphic under perturbation; the return maps on its components are rotations.","feed_headline":"For every number of pieces, stable interval maps are dense","feed_subtitle":"Nudge any piecewise translation slightly and its dynamics becomes a finite union of circle rotations.","key_machinery":"The engine is a transversality theorem, Theorem 4.6 (Coefficients of Linear Dependence), for coefficient vectors in the space $W(r)=\\mathbb{R}^r \\oplus \\mathbb{R}^{r-1}$. To each landing of a discontinuity on another discontinuity, to each return of a branch to a distinguished interval, and to each critical connection, the paper attaches a vector whose coordinates record how many times the orbit visits each partition piece and which discontinuity is hit. Theorem 4.6 states that any linear relation among these dynamically defined vectors forces the coefficients to be equal in a rigid cascade, and Corollary 4.8 turns this into linear independence. That independence lets the authors prescribe small parameter changes that alter one critical return while leaving all other returns and itineraries intact, which is what makes the perturbative approximation of eventually periodic maps by stable maps possible.","core_discovery":"The central discovery is that stability, despite being defined through continuous variation of the non-wandering set, is not a rare property among interval translation maps. The paper establishes Main Theorem I: the set S(r) of stable ITMs on r intervals is dense in ITM(r). Since stable maps are always of finite type, Main Theorem II follows: finite-type maps contain an open and dense subset of ITM(r), while the set of infinite-type maps has empty interior. Density is proved in three stages: eventually periodic maps are dense; a finite-type map is stable exactly when it satisfies the Absence of Critical Connections and Matching conditions; and every eventually periodic map can be perturbed arbitrarily little into a map satisfying those two conditions.","pith_inferences":["The transversality machinery is not tied to the full space ITM(r): the same perturbation-by-independence scheme should make stable maps dense in any rational parameter family that treats the interval endpoints as variable parameters, as the paper itself demonstrates for the Bruin–Troubetzkoy family.","A complete proof covering the omitted boundary cases of Theorem 4.6 would also give a direct route to the paper's Conjecture 8.9, that every eventually periodic map lies in the closure of some stable region, because the current approximation loses control of which stable region is approached as the perturbation shrinks.","Ghost preimages—discontinuities that almost land on each other—are the mechanism by which the non-wandering set can jump upward under perturbation; counting and pruning them could yield combinatorial bounds on the boundary structure of stable regions, bearing on the paper's open questions about convexity and local connectedness."],"forward_implications":["Every ITM on r intervals can be approximated arbitrarily well by a stable map of finite type whose first return maps on the components of its non-wandering set are circle rotations.","The set of infinite-type ITMs has empty interior in ITM(r), giving the topological form of the 1995 measure-theoretic conjecture.","Stability for finite-type ITMs is characterized entirely by two open conditions: no iterate of a discontinuity lands on another discontinuity, and each return map to a component of the non-wandering set has exactly one discontinuity.","If the original measure-theoretic conjecture is true, then almost every ITM actually corresponds to a union of irrational circle rotations.","In the Bruin–Troubetzkoy two-parameter family, the stable maps form a dense subset and the stable regions are exactly the interiors of the coloured triangles shown in the paper."],"supporting_citations":[{"why":"Introduced interval translation maps, constructed the first infinite-type example, and posed the measure-theoretic conjecture whose topological version this paper proves.","marker":"[BK95]"},{"why":"Supplies the characterization of finite type used throughout: an ITM is of finite type exactly when its non-wandering set is a finite union of intervals.","marker":"[ST00]"},{"why":"Developed the renormalization approach and proved measure-zero results for a special three-interval family, which the paper's general method extends and to which Section 8.1 applies the density theorem.","marker":"[BT03]"},{"why":"Established the measure-theoretic conjecture for r=3, the main previously proven full-dimensional case before this paper's all-r topological result.","marker":"[Vol14]"}],"fun_headline_variants":["Stable interval maps are dense","Interval translations: stable maps are dense","Dense stability theorem for interval maps","Stability is dense in interval translation space"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The paper's chain of proofs depends on the full statement of its linear-independence theorem for all boundary cases, but the printed proof only handles the case where boundary points of the distinguished interval land on discontinuities and each relevant point lands on at least two others; the remaining cases are declared analogous.","fun_headline_variants_meta":{"raw":{"variants":["Stable interval maps are dense","Interval translations: stable maps are dense","Dense stability theorem for interval maps","Stability is dense in interval translation space"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00033,"raw_usage":{"total_tokens":1828,"prompt_tokens":925,"completion_tokens":903,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":541,"completion_tokens_details":{"reasoning_tokens":852}},"tokens_in":541,"tokens_out":903,"duration_ms":8615,"temperature":1.0,"reasoning_tokens":852,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T15:18:54.597602+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find an ITM in one of the cases Theorem 4.6 leaves out—say a component interval whose boundary points do not land on discontinuities—whose landing, return, and critical-connection vectors are linearly dependent with non-zero coefficients; the approximation argument would then fail at that point, since every later perturbation is built on that independence.","supporting_citations":[],"review_version":1}