{"id":"4f1c8b33-8c8c-498d-bf19-0fe9bd6f6889","arxiv_id":"2412.07928","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The ARC renormalization yields a natural measure on infinite-type Bruin-Troubetzkoy interval translations, proves unique ergodicity almost everywhere, bounds the Hausdorff dimension of the parameter set between 1.5 and 2, and shows the renormalization cocycle has the Pisot property.","lead":"This paper builds a new renormalization algorithm, the ARC map, for a family of three-interval translation maps and uses it to construct an invariant measure, estimate the fractal dimension of the 'Bruin-Troubetzkoy gasket', and prove a Pisot spectrum property for the algorithm. The results give the first explicit measure and dimension bounds for this family, and reveal a contrast between weak mixing of typical maps and the Pisot property of their renormalization.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 17 is false: for M=A C_A and v=(1,-2,1) in the cone of M, the D-norm ratio is 4/3>1, so Lemma 19 and Theorem 20 (Pisot property) are unsupported as written.","rationale":"Reading in good faith: the renormalization framework and the dimension program are coherent, and much of the paper is a plausible adaptation of known techniques (Fougeron, Jurga, CLL22). The SOSC gap noted by the reader is real but probably repairable: the inverse branches of the R-induction have disjoint cylinders, and the conjugation in Proposition 24 is linear, so it preserves disjointness; this is a missing verification more than a fundamental obstruction. The counterexample to Lemma 17 is a different, more serious issue: it is an internal inconsistency, not a disagreement with consensus. The lemma is used at the heart of Lemma 19, which converts the Birkhoff frequency of the word 112211221 into an exponential contraction for the cocycle norm. If the uncontracted blocks expand, that conversion fails, so the negativity of λ2 is not established. This does not directly disprove the Hausdorff dimension statements in Theorem 4, which use the separate Jurga-based argument, but it invalidates an advertised main theorem and the paper's discussion of the weak-mixing/Pisot phenomenon. The appropriate disposition remains conditional: the authors need either to fix Lemma 17 (the exact norm may still be bounded by a constant, in which case the frequency condition must be rechecked) or to prove Theorem 20 by another method. I therefore keep CONDITIONAL, but for a different and more concrete reason than the reader's.","tokens_in":24462,"tokens_out":23502,"duration_ms":215626,"concrete_test":"Independently recompute the quantity in Lemma 17 for k=1: take M = A C_A, f = (5,3,1) ∈ M R^3_{\\ge0}, and v = (1,-2,1) ∈ f^⊥; verify that ∥v∥_D = 3 and ∥⊤M v∥_D = 4, contradicting the claimed sup = 1. Then, without Lemma 17, rerun the estimate in Lemma 19: if non-paired blocks contribute at least 4/3 and occur at positive frequency, the upper bound becomes 2(4/3)^m(4/5)^{#I_m} with m ∼ n·ν(C_A,C_B) and #I_m ∼ (1/8)n·µ([112211221]); check whether this exponent can still be negative, and if not, Theorem 20 has no proof.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Lemma 17 asserts that ∥⊤(A^k C_A)|_{R^3_{\\ge0}}\\|_D = ∥⊤(B^k C_B)|_{R^3_{\\ge0}}\\|_D = 1 for every k≥0. This equality is already false for k=1. Let M = A C_A; then M = [[2,1,1],[0,1,0],[1,0,1]] and ⊤M = [[2,0,1],[1,1,0],[1,0,1]]. The vector f = (5,3,1) lies in M R^3_{\\ge0}, since it equals (2,0,1)+3(1,1,0). The vector v = (1,-2,1) lies in f^⊥ because 5·1+3·(−2)+1·1 = 0. Direct computation gives ∥v∥_D = 3 and ∥⊤M v∥_D = 4, so the restricted norm is at least 4/3 > 1. The error in the proof is the assertion 'min v ≤ min v′ ≤ 0 ≤ max v′ ≤ max v' for v′ = (v1, v2−v3, 0); here v′ = (1,−3,0), whose minimum is −3 < min v = −2. Lemma 19 uses Lemma 17 to bound every non-paired block A^k C_A / B^k C_B by 1. With expansion factors > 1, the bound ∥⊤M[0,nm)|f^⊥∥_∞ ≤ 2 (4/5)^{#I_m} no longer follows, and the contraction in Theorem 20 may be overwhelmed by the positive frequency of these blocks. Hence the proof of the negative second Lyapunov exponent, and with it Theorem 5 (Pisot property), is not valid as written.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces an induction/renormalization procedure R for the Bruin–Troubetzkoy family of 3-interval translation maps, interprets it as a Markovian multidimensional continued fraction algorithm (ARC), and derives three main results: (i) a renormalization-invariant measure μ supported on the infinite-type parameter set G with respect to which T_{α,β} is uniquely ergodic almost surely (Theorem 3); (ii) Hausdorff dimension bounds 1.5 ≤ dim_H G < 2 and equality with the affinity dimension (Theorem 4); and (iii) an almost-sure Pisot Lyapunov spectrum for the ARC cocycle (Theorem 5). The proofs combine Fougeron's simplicial-system criteria, the CLL22 Lyapunov-exponent method, and Jurga's results on the Rauzy gasket.","tokens_in":24914,"tokens_out":9072,"duration_ms":85158,"significance":"The significance is potentially high. The paper offers the first natural renormalization-invariant measure for infinite-type Bruin–Troubetzkoy interval translation maps, connects the parameter set to a gasket analogous to the Rauzy gasket, and exhibits a continued-fraction algorithm with the Pisot property whose associated ITMs are conjecturally weak mixing. The manuscript is well organized and contains explicit matrix computations that are in principle checkable. These strengths are real. However, one central lemma is false and a key open-set condition is asserted without proof, so the present form does not establish Theorems 4 and 5.","major_comments":[{"comment":"Lemma 17 is false as stated. For k=1, write M = A C_A. Then ⊤M = [[2,0,1],[1,1,0],[1,0,1]]. Take f = (1,2,3) ∈ R^3_{\\ge 0} and v = (1,-2,1) ∈ f^\\perp. A direct computation gives ∥v∥_D = 3 and ∥⊤M v∥_D = 4, so the norm restricted to f^\\perp is at least 4/3 > 1, contradicting the claimed value 1. The proof's auxiliary vector v' = (v_1, v_2 - v_3, 0) = (1,-3,0) violates the asserted inequality min v ≤ min v' because min v = -2 and min v' = -3. Since Lemma 19 uses Lemma 17 to discard all non-paired blocks A^k C_A and B^k C_B, and Theorem 20 derives the negative second Lyapunov exponent from Lemma 19, the proof of Theorem 5 is unsupported as written. A corrected argument must either prove a weakened norm bound that still yields contraction, or replace this step entirely.","section":"Section 3.2, Lemma 17"},{"comment":"The proof asserts that, after conjugation, the matrices Γ_N 'satisfy the SOSC', but no open set U or verification of f_i(U) ∩ f_j(U) = ∅ is actually supplied. Theorem 21 (Jurga's Theorem 1.3), which is the mechanism for concluding dim_H K_X = min{s_X, 2}, explicitly requires the strong open set condition. Balancedness (Proposition 24) and Zariski density (Proposition 27) do not by themselves imply disjointness of images. Without a proof of SOSC, the chain sup_N dim_H K_{Γ_N} = sup_N s_{Γ_N} = s_Γ = s_A is not justified, and therefore the equality dim_H G = s_A in Theorem 4 is not established. Please provide a proof or a precise reference for the SOSC claim for the conjugated systems.","section":"Section 4.3, proof of Theorem 22"},{"comment":"The proof of Lemma 19 begins with the unproved assertion that μ assigns positive measure to every cylinder, in particular μ([112211221]) > 0. This positivity is load-bearing: without it, the exponent in the final bound may be non-positive and the inequality cannot imply λ_2 < 0. Positivity on all cylinders does not follow immediately from the support statement in Theorem 3; an argument is needed, for example that each cylinder intersects the parameter simplex in a nonempty open set and that μ has full support. Please state and prove the required positivity property.","section":"Section 3.2, Lemma 19"}],"minor_comments":[{"comment":"The name 'Bruin-Troubetzoy' is a typo for 'Bruin-Troubetzkoy'.","section":"Section 2.2, Lemma 7"},{"comment":"The word 'ineuquality' should be 'inequality'; also the notation n_m versus n in the Birkhoff-sum estimate should be made consistent, since the paragraph uses both without explicitly relating the two.","section":"Section 3.2, proof of Lemma 19"},{"comment":"The phrase 'identity the alphabet A with the set of matrices' should be 'identify the alphabet A with the set of matrices'.","section":"Section 4.2"},{"comment":"The notation ∥M|_{f^⊥}∥_D is used before the restricted norm is defined; the convention introduced in equation (5) should be stated immediately after Lemma 16 for readability.","section":"Section 3.2, Lemma 16"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe ARC renormalization for Bruin-Troubetzkoy ITMs is a genuinely new framework, and the paper does well to import Fougeron's simplicial system machinery and Jurga's dimension results into this setting. The construction of an invariant measure and the unique ergodicity statement (Theorem 3) are plausible and follow a known template; the upper bound on dim H G also looks fine. The dimension lower bound and the Pisot property are where the problems sit.\n\nThe biggest issue is concrete: Lemma 17 is false. For k=1, take f = (5,3,1), which lies on the cone generated by A C_A, and v = (1,-2,1) in f^⊥. Then ||v||_D = 3 and ||⊤(A C_A) v||_D = ||(3,-1,2)||_D = 4, so the norm of the restriction is at least 4/3, not 1. The proof's claim that min v ≤ min v' for v' = (v1, v2−v3, 0) fails: for this v, v' = (1,−3,0) has min −3, which is less than min v = −2. Since Lemma 19 uses Lemma 17 to bound every non-paired block by 1, the contraction estimate in Theorem 20 collapses. The negative Lyapunov exponent, and hence Theorem 5 (Pisot property), is not proven as written. The error is load-bearing, not cosmetic.\n\nA second gap: Theorem 22 asserts, without verification, that the conjugated IFS from Γ_N satisfies the strong open set condition. Jurga's theorem requires SOSC, and the paper neither constructs the open set nor checks disjointness. So the equality dim_H G = s_A is unsupported, although the bounds 1.5 ≤ dim_H G < 2 might still hold through other arguments.\n\nA minor point: Lemma 19 asserts that µ assigns positive mass to every cylinder, which is not automatic and is not shown.\n\nWhat is good: the ARC algorithm itself, the connection to the Rauzy gasket literature, and the measure construction are genuinely novel. The paper is clearly written and the external theorems are used appropriately, modulo the unchecked hypotheses. The identified phenomenon—Pisot renormalization alongside typical weak mixing—is interesting, but it now rests on a broken proof.\n\nThis paper deserves a serious referee, but the referee should require a corrected Lemma 17 and an actual SOSC proof. As it stands, the advertised Pisot theorem and the affinity dimension equality are unproven.\n\nRecommendation: send to peer review, expecting major revision.","headline":"A promising renormalization framework for Bruin-Troubetzkoy ITMs, but a false lemma in the Pisot proof and an unverified SOSC leave two of the three main theorems unproven as written.","tokens_in":25411,"tokens_out":6271,"would_cite":false,"duration_ms":55250,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["37E05","37A05","37A44","11J70"],"pacs":[],"model":"deepseek-v4-flash","headline":"A new Rauzy-style induction for Bruin-Troubetzkoy interval translation mappings produces a renormalization-invariant measure and shows the infinite-type parameter set has Hausdorff dimension between 1.5 and 2, matching its affinity…","keywords":["interval translation mappings","renormalization","multidimensional continued fractions","Pisot property","Hausdorff dimension","affinity dimension","unique ergodicity","Bruin-Troubetzkoy gasket"],"falsifier":"For a fixed truncation level, write down the finitely many projective maps in the conjugated iterated function system and check whether their images on the open simplex are pairwise disjoint; a single overlap with nonempty interior violates SOSC and would break the application of the theorem giving dimension equals affinity dimension. Alternatively, compute the pressure function for the truncated semigroup and test whether its unique root approaches the claimed affinity dimension as the truncation level grows.","tokens_in":24278,"feed_emoji":"📐","tokens_out":8187,"duration_ms":77247,"temperature":0.7,"pith_summary":"The paper studies the two-parameter family of three-interval translation maps introduced by Bruin and Troubetzkoy, maps of the unit interval made of three translations whose images may overlap and leave holes. It proposes a renormalization scheme, an induction on the interval similar in spirit to Rauzy induction for interval exchanges, and views it as a multidimensional continued fraction algorithm. The central aim is to show that the parameter set giving infinite-type (Cantor-attractor) maps carries a natural renormalization-invariant measure, that almost every such map is uniquely ergodic, and that this parameter set is a fractal with Hausdorff dimension between 1.5 and 2 equal to its affinity dimension. If correct, this turns a family previously understood through a Gauss map and symbolic substitutions into a self-affine gasket controlled by a matrix cocycle, and it exhibits a new combination: the renormalization algorithm has the Pisot property almost surely while the underlying maps are expected to be typically weak mixing.","feed_headline":"Infinite-type ITM parameters form a fractal of dimension 1.5 to 2","feed_subtitle":"Rauzy-style induction yields an invariant measure; almost every such interval translation map is uniquely ergodic.","key_machinery":"The load-bearing object is the induction on the simplex of interval lengths , with Cases 1 and 3 coded by the matrices , , and their symmetric analogues , ; iterating the induction gives a Markovian multidimensional continued fraction algorithm, named the Arnoux-Rauzy-Cassaigne (ARC) algorithm, whose projectivized action is a simplicial system. The invariant measure is obtained as the projection of the measure of maximal entropy on the natural suspension of this simplicial system. For the dimension result, the infinite-type set is realized as the attractor of an iterated function system generated by the semigroup built from these matrices, and the affinity dimension is the critical exponent of the singular-value zeta function of that semigroup.","core_discovery":"The central discovery is that the Bruin-Troubetzkoy gasket, the set of parameters for which the interval translation map is of infinite type, is a self-affine fractal of the same general kind as the Rauzy gasket, and its Hausdorff dimension coincides with the affinity dimension of an associated semigroup of positive matrices. The authors construct an induction with four basic moves, encode it as a simplicial system, and use the simplicial-system machinery to produce a renormalization-invariant measure fully supported on this parameter set; for almost every such parameter the interval translation map is uniquely ergodic. They prove that the Hausdorff dimension lies between 1.5 and 2 and equals the affinity dimension, and they show that the induced multidimensional continued fraction algorithm has an almost everywhere negative second Lyapunov exponent, i.e., the Pisot property. Together these results upgrade earlier qualitative statements about this family into a measure-theoretic and dimension-theoretic picture.","pith_inferences":["If the weak-mixing conjecture is confirmed, Bruin-Troubetzkoy interval translation maps would be a natural family where a Pisot renormalization cocycle coexists with almost sure weak mixing of the underlying systems, showing that the Pisot property does not automatically force spectral rigidity.","The same induction should extend to the d-branch Bruin-Troubetzkoy family studied by Bruin, producing analogous gaskets whose dimension bounds are governed by the corresponding matrix semigroup; the paper does not treat that generalization.","Because the ARC algorithm comes with an explicit invariant measure and an almost sure Pisot property, it is a candidate for simultaneous Diophantine approximation of pairs of parameters; testing its convergence rates numerically would be a direct check of whether the Pisot property is useful there.","One could numerically approximate the Hausdorff dimension by sampling long admissible words in the semigroup and estimating the singular-value pressure; agreement with the affinity dimension would corroborate the open-set assumption, while a mismatch would localize where the proof needs repair."],"forward_implications":["There is a renormalization-invariant measure with support exactly equal to the infinite-type parameter set, and almost every Bruin-Troubetzkoy interval translation map with respect to it is uniquely ergodic.","The infinite-type parameter set has Hausdorff dimension at least 1.5 and strictly less than 2, so it is a measure-zero but dimension-rich fractal set.","Hausdorff dimension equals the affinity dimension, so the dimension is determined by the singular-value function of the matrix semigroup and can be studied through pressure functions.","The ARC multidimensional continued fraction algorithm has an almost everywhere negative second Lyapunov exponent, giving the Pisot property for the renormalization cocycle.","The paper sets up the conjecture that almost every infinite-type Bruin-Troubetzkoy interval translation map is weakly mixing, which would contrast with the almost sure Pisot property of the renormalization algorithm."],"supporting_citations":[{"why":"Supplies the simplicial-system criterion used to construct the invariant measure and the upper bound on Hausdorff dimension.","marker":"[Fou20]"},{"why":"Defines the Bruin-Troubetzkoy family and its Gauss map; proves the zero-Lebesgue-measure and dense-G-delta results that this paper refines.","marker":"[BT03]"},{"why":"Provides the template theorem that SOSC plus Zariski density yields the Hausdorff dimension as the affinity dimension, used in proving the equality.","marker":"[Jur23]"},{"why":"Supplies the lower-bound strategy for the dimension via stationary measures and limit sets, used to obtain the 1.5 lower bound.","marker":"[Jia+24]"},{"why":"Provides the norm lemmas and Pisot-spectrum techniques used to prove the almost sure Pisot property of the ARC algorithm.","marker":"[CLL22]"},{"why":"Provides the standard argument converting renormalization-invariant measures into unique ergodicity of the original interval translation maps.","marker":"[Vee82]"},{"why":"Supplies the geometric covering lemma used in the lower-bound dimension proof.","marker":"[PS23]"},{"why":"Supplies the notion of special acceleration and related simplicial-system results used in the proof of the invariant-measure theorem.","marker":"[FS21]"}],"fun_headline_variants":["Bruin-Troubetzkoy gasket is self-affine with Hausdorff dimension = affinity","Almost every infinite-type ITM uniquely ergodic via new renormalization","Renormalization reveals fractal set of dimension 1.5-2 for ITMs","Pisot property almost always: new continued fraction from renormalization"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The equality between Hausdorff dimension and affinity dimension rests on the strong open set condition (SOSC) for the conjugated iterated function system generated by the truncated matrix sets, which the paper asserts but does not verify by constructing the required open set.","fun_headline_variants_meta":{"raw":{"variants":["Bruin-Troubetzkoy gasket is self-affine with Hausdorff dimension = affinity","Almost every infinite-type ITM uniquely ergodic via new renormalization","Renormalization reveals fractal set of dimension 1.5-2 for ITMs","Pisot property almost always: new continued fraction from renormalization"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001221,"raw_usage":{"total_tokens":4995,"prompt_tokens":892,"completion_tokens":4103,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":508,"completion_tokens_details":{"reasoning_tokens":4015}},"tokens_in":508,"tokens_out":4103,"duration_ms":28685,"temperature":1.0,"reasoning_tokens":4015,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T18:24:46.311072+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"For a fixed truncation level, write down the finitely many projective maps in the conjugated iterated function system and check whether their images on the open simplex are pairwise disjoint; a single overlap with nonempty interior violates SOSC and would break the application of the theorem giving dimension equals affinity dimension. Alternatively, compute the pressure function for the truncated semigroup and test whether its unique root approaches the claimed affinity dimension as the truncation level grows.","supporting_citations":[],"review_version":1}