{"id":"10fdd68e-6f6d-47a6-ad56-d55a1a3471e0","arxiv_id":"1908.05954","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":1.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A survey of S-adic sequences and their correspondence, under natural conditions, with rotations on higher-dimensional tori, including metric genericity results.","lead":"An expert survey explains how sequences of letters, known as S-adic sequences, can represent rotations on higher-dimensional tori, generalizing the classical link between Sturmian sequences and circle rotations. It covers recent theorems that give measurable conjugacies under natural conditions and shows when such representations are typical.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The measurable conjugacy in Theorem 3.9.4 rests entirely on the extra C1 tiling condition, and Theorem 3.9.5 as stated does not show this condition is generic; the metric claim that counterexamples are rare is thus unsupported by the presented argument.","rationale":"The reader's weakest_assumption correctly identifies the C1 tiling / geometric coincidence condition as load-bearing for Theorem 3.9.4. My reading agrees with that identification: the tiling is not derived from the other assumptions, and the survey itself presents it as a separate, hard-to-verify combinatorial condition. My concern goes one step further: the metric version, Theorem 3.9.5, keeps the tiling condition as a hypothesis in part (3), and the proof sketch in Section 3.9.4 does not show this hypothesis is ν-a.e. satisfied. Since the abstract and introduction emphasize a metric theory showing that counterexamples are rare, this is the place where the presented argument is least secure. I do not think this changes the overall verdict: the survey is honest about the conditional nature of the conjugacy result, and the classical parts are well supported. The right verdict remains CONDITIONAL, with the condition being precisely the tiling/genericity gap, so UNCHANGED is appropriate.","tokens_in":64212,"tokens_out":5532,"duration_ms":59133,"concrete_test":"Inspect the full proof of Theorem 3.9.5 in [52] and locate the lemma or argument that establishes the C1 tiling / geometric coincidence condition for ν-almost every σ under the Pisot condition. If no such step exists, the metric claim is unproven. Complementarily, run the effective criterion of Proposition 3.8.8(iv) on a random sample of recurrent sequences from an ergodic S-adic shift satisfying the Pisot condition (e.g., Brun or Arnoux-Rauzy substitutions) and estimate the ν-measure of sequences for which C1 fails to tile; a positive-measure failure set would show the 'if' clause in Theorem 3.9.5(3) is essential and the introduction's 'rare' statement needs a quantitative formulation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central result of the chapter is Theorem 3.9.4, which asserts that a primitive, algebraically irreducible, balanced, recurrent S-adic system is measurably conjugate to a rotation on T^{d-1}. The theorem's hypothesis is not just these recurrence/balance/irreducibility conditions: it also requires that the collection C1 of Rauzy-fractal translates tile 1^⊥. Proposition 3.8.8 shows this tiling is equivalent to the geometric coincidence condition, and Section 3.8.2 explicitly introduces geometric coincidence as an additional, non-derived input. Nothing in Sections 3.5-3.7 proves that primitivity, recurrence, finite balance, and algebraic irreducibility imply geometric coincidence; indeed Section 3.8.3 is devoted to the nontrivial problem of checking it in examples. In the proof of Theorem 3.9.4, the tiling condition is used at the decisive point where R is identified with a fundamental domain for the lattice Λ = 1^⊥ ∩ Z^d; without it, the representation map is m-to-1 with m the covering degree, and the induced map is not a rotation on T^{d-1}. The metric Theorem 3.9.5 still contains this condition only as the hypothesis of part (3): 'If the collection C1 associated with σ forms a tiling ...'. The proof sketch in Section 3.9.4 establishes genericity of primitivity/recurrence, balance, and algebraic irreducibility, but no displayed lemma proves that C1 tiling (or geometric coincidence) holds for ν-almost every σ. Consequently, the abstract's claim that the metric theory makes Cassaigne-Ferenczi-Zamboni-type counterexamples 'rare' is stronger than the presented argument supports unless a missing tiling-genericity step is supplied. Remark 3.3.8, where the author admits an inability to verify a published proof, reinforces the need for caution when relying on unstated verification steps.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This survey chapter presents the S-adic approach to Rauzy's program, aiming to bridge Sturmian-type combinatorics, generalized continued fraction algorithms, and rotations on higher-dimensional tori. The first part gives a largely self-contained exposition of the classical correspondence between Sturmian sequences, the continued fraction/Gauss map, and irrational rotations, including the natural-extension picture via the geodesic and scenery flows on SL_2(Z)\\SL_2(R). The second part develops the general machinery: S-adic systems, Rauzy fractals and their subtiles, primitivity, recurrence, balance, algebraic irreducibility, strong convergence, coincidence conditions, and tiling properties. The chapter then states and sketches the proofs of two main results. Theorem 3.9.4 asserts that, for a primitive and algebraically irreducible sequence of unimodular substitutions satisfying a recurrence/balance hypothesis and a tiling condition on C1, the S-adic shift is measurably conjugate to a rotation on T^{d-1}. Theorem 3.9.5 is a metric version: under an ergodic Pisot condition on the associated linear cocycle, almost every directive sequence yields minimal, uniquely ergodic, well-behaved Rauzy fractals, and conditionally on the C1 tiling property one obtains the measurable conjugacy to a torus rotation. The chapter closes with applications to Arnoux-Rauzy and Brun substitution systems.","tokens_in":64549,"tokens_out":9296,"duration_ms":93125,"significance":"If the results reported from Berthé–Steiner–Thuswaldner [52] are correct, this is a valuable survey that organizes a large body of material and gives a coherent roadmap for Rauzy's program. Its strengths are the detailed proofs of the classical Sturmian theory, the precise references for the newer results, and the honest discussion of several technical points, including the author's explicit remark in Remark 3.3.8 that one stronger statement from [63] could not be verified. The chapter also does a good job of explaining why primitivity, recurrence, balance, and algebraic irreducibility are natural hypotheses. The principal weakness is that the advertised metric genericity claim goes beyond what is actually proved or stated in the chapter: Theorem 3.9.5(3) is conditional on the C1 tiling condition, and no argument is supplied showing that this condition is generic.","major_comments":[{"comment":"Theorem 3.9.5(3) is stated conditionally on the collection C1 forming a tiling of 1^⊥, and the proof in Section 3.9.4 establishes genericity only for primitivity and recurrence (Proposition 3.9.6), finite balance (Lemma 3.9.7), and algebraic irreducibility (Lemma 3.9.8). No lemma or displayed argument in Section 3.9.4 shows that the tiling condition, equivalently geometric coincidence by Proposition 3.8.8, holds for ν-almost every σ. Since the abstract and Section 3.1 advertise a metric theory showing that counterexamples of the Cassaigne–Ferenczi–Zamboni type are rare, this claim is not supported by the chapter as written; the author should either supply the missing genericity statement with a precise reference to [52] or rephrase the abstract and introduction to present Theorem 3.9.5(3) as a conditional result.","section":"§3.9.3–3.9.4 and Abstract"},{"comment":"The tiling condition is not a harmless technicality: the proof of the conjugacy uses it at the decisive step where the representation map is identified with a rotation on 1^⊥/Λ, and the text explicitly notes that without a tiling the map φ is m-to-1 with m equal to the covering degree of C1. This should be stated prominently when the theorem is announced, and Section 3.8 should make clear that primitivity, recurrence, balance, and algebraic irreducibility do not by themselves imply geometric coincidence; Section 3.8.3 indeed describes checking this condition as substantial work. The current wording of the abstract, namely \"under certain natural conditions\", obscures that the tiling hypothesis is an additional, independently checkable condition.","section":"§3.9.2 / Theorem 3.9.4"}],"minor_comments":[{"comment":"In the definition of C^{(k)}_w and in Proposition 3.7.6, the translation term is written as π_{u,w}x, but consistency with (3.43)–(3.45) requires π^{(k)}_{u,w}x, since R^{(k)}_w(i) lives in (w^{(k)})^⊥ and the sets being translated must lie in the same hyperplane; as written, the left-hand side is not even a subset of a single hyperplane.","section":"§3.7.1"},{"comment":"Definition 3.9.1 contains the typo \"Defintion\" in the citation to [126]; it should read \"Definition\".","section":"§3.9.1"},{"comment":"The remark honestly records that the author could not verify the unbounded-fundamental-domain version of [63, Corollary 2.6], but the survey would benefit from repeating this limitation when Corollary 3.3.7 is cited, so that readers do not attribute the stronger statement to this chapter.","section":"§3.3.2 / Remark 3.3.8"},{"comment":"In the proof sketch of Theorem 3.9.4, the sentence \"E is a surjective piecewise isometry. Therefore, E is bijective\" is too quick: surjectivity alone does not imply injectivity for piecewise isometries, and the argument should explicitly mention that the tiling hypothesis and measure disjointness of the subtiles are what make the map bijective almost everywhere.","section":"§3.9.2"}],"recommendation":"major_revision","confidential_remarks":"The main concern is that the chapter's advertised metric genericity statement may exceed what is demonstrated in the text. If [52] contains a full-measure tiling theorem, the author should cite it explicitly at Theorem 3.9.5(3); if not, the abstract and introduction need to be weakened. The heavy reliance on [52], co-authored by the chapter's author, for all central theorems is not by itself a defect of a survey, but it does mean that an editor may wish to ensure that the precise statements quoted from [52] have been checked by a referee of the research paper."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The main value of this chapter is as a map. It takes the reader from Morse-Hedlund and Coven-Hedlund through Rauzy induction and Sturmian codings to the recent Berthé-Steiner-Thuswaldner theorem that, under enough hypotheses, an S-adic system is measurably conjugate to a torus rotation. The classical parts are genuinely well done and mostly self-contained; the modern parts are precise about statements and give useful proof sketches with references. It is a survey and does not claim new theorems, so the low 'novelty' score is beside the point. For someone outside the area, this is probably the clearest available route into Rauzy's program.\n\nThe soft spots are real but not disqualifying. The central Theorem 3.9.4 is quoted from the author's own paper [52]; the chapter gives a sketch, not a proof, so the reader's confidence has to rest on [52]. That is normal for a survey, but it matters here because the theorem carries a load-bearing extra hypothesis: the collection C1 must tile the hyperplane 1^⊥, which is equivalent to a geometric coincidence condition. Nothing in the chapters on primitivity, recurrence, balance, or algebraic irreducibility implies that tiling. Theorem 3.9.5, the metric statement, establishes genericity for minimality, unique ergodicity, and the topological regularity of the Rauzy fractals, but it only gives the conjugacy in part (3) conditional on the same tiling. The abstract's claim that Cassaigne-Ferenczi-Zamboni-type counterexamples are 'rare' is therefore stronger than the presented argument demonstrates, unless the missing tiling-genericity step lives in [52]. The author should either add that step or soften the abstract.\n\nRemark 3.3.8 is also worth noting: the author admits he could not verify the published proof of the unbounded-fundamental-domain version of the CFZ counterexample. That is honest, and it is a good reason to keep that statement as a remark rather than a theorem.\n\nWho should read it: graduate students and researchers in symbolic dynamics, Diophantine approximation, or tiling theory who want the landscape. It deserves serious peer review; a careful referee should ask for the metric-claim fix, but the chapter itself is a solid, useful survey.","headline":"A genuinely useful survey of Rauzy's program whose central theorem is honestly quoted from the author's own paper; the abstract's metric 'rare counterexamples' claim, however, outruns the theorem as stated.","tokens_in":65116,"tokens_out":2913,"would_cite":true,"duration_ms":31883,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["37B10","11K50","37A30"],"pacs":[],"model":"deepseek-v4-flash","headline":"S-adic sequences generalize Sturmian coding to rotations on higher-dimensional tori.","keywords":["S-adic sequences","Rauzy fractals","torus rotations","pure discrete spectrum","generalized continued fractions","Rauzy induction","Arnoux-Rauzy sequences","Brun substitutions"],"falsifier":"Run the theorem's own construction on any sequence satisfying all its hypotheses: build the representation map $\\phi:X_\\sigma\\to R$ from nested subtiles and check whether it is injective off a set of measure zero. If two positive-measure sets of sequences collapse to the same Rauzy-fractal point, or if the shift has a measurable eigenfunction not coming from the torus rotation, the claimed conjugacy fails; a numerical implementation for a Brun or Arnoux-Rauzy example would be a direct check.","tokens_in":64016,"feed_emoji":"🔁","tokens_out":14525,"duration_ms":129503,"temperature":0.7,"pith_summary":"This paper surveys and proves a higher-dimensional generalization of the classical correspondence between Sturmian sequences, continued fractions, and irrational rotations of the circle. Its central theorem states that under natural hypotheses—primitivity, algebraic irreducibility, recurrence, balance, and a tiling condition on the associated Rauzy fractal—the shift dynamical system generated by an S-adic sequence is measurably conjugate to a rotation on the $(d-1)$-dimensional torus, so the system has purely discrete spectrum. The proof shows concretely that each sequence in the system is a natural coding of that rotation and that the fractal subtiles are bounded remainder sets. A metric counterpart, proved under a Pisot condition on Lyapunov exponents, shows these good properties hold for almost every substitution sequence in the shift, so the known imbalanced and weakly mixing counterexamples are rare rather than generic.","feed_headline":"S-adic systems are rotations on higher-dimensional tori","feed_subtitle":"Generalizing Sturmian sequences, generic S-adic systems are measurably conjugate to torus rotations.","key_machinery":"One load-bearing object is the S-adic Rauzy fractal: for a sequence $\\sigma$ with letter-frequency vector $u$, the set $R_w$ is the closure of the projection, along $u$ onto the hyperplane $w^\\perp$, of the abelianized prefixes of all limit sequences; the subtile $R_w(i)$ keeps only prefixes ending in the letter $i$. These sets admit a subdivision governed by the set equation, expressed through the dual substitution $E_1^*(\\sigma)$ acting on discrete hyperplanes $\\Gamma(w)$, and the diameter of the level-$n$ pieces tends to zero under the hypotheses of the theorem. The collection $\\mathcal{C}_1=\\{\\pi_{u,1}x+R(i):[x,i]\\in\\Gamma(1)\\}$ is the decisive object: when it tiles $1^\\perp$, which Proposition 3.8.8 equates with the geometric coincidence condition, the fractal $R$ is a fundamental domain of the lattice $1^\\perp\\cap\\mathbb{Z}^d$, the piecewise translation exchange on the subtiles becomes a rotation on the torus $1^\\perp/\\Lambda\\cong T^{d-1}$, and the representation map $X_\\sigma\\to R$ becomes a measurable isomorphism. Balance buys compactness of the tiles, algebraic irreducibility buys rational independence of $u$ and strong convergence, and the Pisot Lyapunov condition makes the whole package generic in the metric theorem.","core_discovery":"At the center is Theorem 3.9.4: for a finite set of unimodular substitutions and a sequence $\\sigma$ that is primitive and algebraically irreducible, whose shifted languages are eventually $C$-balanced and whose blocks repeat (the recurrence condition), if the collection $\\mathcal{C}_1$ of translates of the S-adic Rauzy fractal tiles the hyperplane $1^\\perp$, then the S-adic shift $(X_\\sigma,\\Sigma,\\mu)$ is measurably conjugate to a rotation on $T^{d-1}$. Consequently the shift has purely discrete spectrum; every element of $X_\\sigma$ is a natural coding of that rotation with respect to the partition into fractal subtiles $R(1),\\ldots,R(d)$; and each subtile is a bounded remainder set. The chapter also establishes the metric version, Theorem 3.9.5: inside an ergodic shift over a finite substitution set satisfying the Pisot condition $\\vartheta_1>0>\\vartheta_2\\geq\\cdots\\geq\\vartheta_d$, almost every $\\sigma$ gives a minimal, uniquely ergodic system whose Rauzy fractals have the required regularity, and those that satisfy the tiling condition are conjugate to torus rotations. The upshot is that the classical triangle linking Sturmian sequences, continued fractions, and circle rotations is realized in dimension $d-1$, with generalized continued fraction algorithms encoded by the substitution sequences.","pith_inferences":["If the theorem is correct, the practical question for any concrete substitution family is whether geometric coincidence holds, and Proposition 3.8.8 turns that into a finite check; one could automate this for recursively generated families and map which S-adic systems are rotation codings.","The metric theorem suggests a dichotomy that can be tested numerically: within an ergodic Pisot substitution shift, either almost every sequence satisfies the tiling condition or the exceptional set has positive measure, and random sampling over the Brun alphabet would distinguish the two regimes.","The construction may extend to a coding of higher-rank diagonal actions such as the Weyl chamber flow, in the way the classical continued fraction codes the geodesic flow on the modular surface; the Rauzy-fractal tilings would then serve as non-stationary Markov partitions.","The chapter's separation of balance from the tiling condition indicates that imbalance and failure of geometric coincidence are the two obstructions that stand between a general S-adic system and a rotation model; other dynamical pathologies, such as weak mixing, should be traceable to one of them."],"forward_implications":["Every S-adic system satisfying the hypotheses of Theorem 3.9.4 has purely discrete spectrum, so it is as far from weakly mixing as a measure-preserving system can be.","Elements of $X_\\sigma$ are natural codings of an explicit rotation on $T^{d-1}$, so the substitution coding sequence carries the Diophantine data of the rotation vector.","The subtiles $R(i)$ are bounded remainder sets: every word frequency in the S-adic sequence deviates from its limit by a uniformly bounded error.","Under the Pisot condition on Lyapunov exponents, almost every substitution sequence produces balanced, primitive, recurrent, algebraically irreducible S-adic systems; the tiling condition is the only additional hypothesis needed for conjugacy.","The framework applies to recurrent Arnoux-Rauzy and Brun substitution sequences, giving explicit higher-dimensional analogues of the Sturmian–continued-fraction–rotation correspondence."],"supporting_citations":[{"why":"Supplies the main conjugacy theorem (Theorem 3.9.4), the metric theorem (Theorem 3.9.5), and the set-equation and coincidence machinery the chapter presents.","marker":"[52]"},{"why":"Introduces the Tribonacci substitution and the first Rauzy fractal, the prototype whose tiling gives conjugacy to a two-dimensional torus.","marker":"[114]"},{"why":"Defines Arnoux-Rauzy sequences and the S-adic coding and induction that the higher-dimensional theory generalizes.","marker":"[22]"},{"why":"Constructs imbalanced Arnoux-Rauzy sequences, the counterexamples showing the correspondence requires extra hypotheses beyond the classical ones.","marker":"[63]"},{"why":"Builds weakly mixing Arnoux-Rauzy systems, the exceptional behavior that the metric theorem shows to be rare.","marker":"[62]"},{"why":"Proves how the dual substitution maps discrete hyperplanes, a key step in the set equation and in the coincidence conditions.","marker":"[79]"},{"why":"Develops the annulus method for verifying geometric finiteness for Arnoux-Rauzy sequences, making the tiling condition checkable in examples.","marker":"[48]"},{"why":"Provides the weak primitivity and uniform-frequency framework that yields minimality and unique ergodicity for S-adic systems.","marker":"[44]"}],"fun_headline_variants":["S-adic systems: new bridge from dynamics to torus rotations","S-adic sequences reveal rotations on higher-dimensional tori","Generalized Sturmian sequences: S-adic shifts conjugate to rotations","S-adic coding gives measurable conjugacy to torus rotations","Rauzy's program realized: S-adic shifts rotate on tori"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the translates in $\\mathcal{C}_1$ tile the hyperplane $1^\\perp$ exactly once—equivalently, the geometric coincidence condition—since primitivity, recurrence, balance, and algebraic irreducibility do not by themselves force this tiling.","fun_headline_variants_meta":{"raw":{"variants":["S-adic systems: new bridge from dynamics to torus rotations","S-adic sequences reveal rotations on higher-dimensional tori","Generalized Sturmian sequences: S-adic shifts conjugate to rotations","S-adic coding gives measurable conjugacy to torus rotations","Rauzy's program realized: S-adic shifts rotate on tori"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000527,"raw_usage":{"total_tokens":2657,"prompt_tokens":1172,"completion_tokens":1485,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":788,"completion_tokens_details":{"reasoning_tokens":1404}},"tokens_in":788,"tokens_out":1485,"duration_ms":12874,"temperature":1.0,"reasoning_tokens":1404,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:59:36.410244+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the theorem's own construction on any sequence satisfying all its hypotheses: build the representation map $\\phi:X_\\sigma\\to R$ from nested subtiles and check whether it is injective off a set of measure zero. If two positive-measure sets of sequences collapse to the same Rauzy-fractal point, or if the shift has a measurable eigenfunction not coming from the torus rotation, the claimed conjugacy fails; a numerical implementation for a Brun or Arnoux-Rauzy example would be a direct check.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the main conjugacy theorem (Theorem 3.9.4), the metric theorem (Theorem 3.9.5), and the set-equation and coincidence machinery the chapter presents."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the Tribonacci substitution and the first Rauzy fractal, the prototype whose tiling gives conjugacy to a two-dimensional torus."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Constructs imbalanced Arnoux-Rauzy sequences, the counterexamples showing the correspondence requires extra hypotheses beyond the classical ones."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Builds weakly mixing Arnoux-Rauzy systems, the exceptional behavior that the metric theorem shows to be rare."},{"cited_title":"Internat","cited_arxiv_id":null,"evidence_quote":"Proves how the dual substitution maps discrete hyperplanes, a key step in the set equation and in the coincidence conditions."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Develops the annulus method for verifying geometric finiteness for Arnoux-Rauzy sequences, making the tiling condition checkable in examples."}],"review_version":1}