{"id":"0948b519-d174-4eae-943d-90a3a429849b","arxiv_id":"2507.09775","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For certain Veech surfaces, the paper shows that the Teichmüller geodesic flow pushes twist tori to dense sets, and under an algebraic monodromy condition every weak-* limit is fully supported.","lead":"This math paper proves that, for certain symmetric translation surfaces, the expanding images of a family of tori become dense in the expected moduli space, and in some cases all limit distributions have full support. It is a partial step toward conjectures of Forni and Mirzakhani about equidistribution of horocycle and earthquake flows.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Tremor-scaling mismatch in the proof of Proposition 5.1: Lemma 7.2 controls tremors by r·(unit vector), while Tremβ(t,s,r) uses r·β(t,s); the proof needs the extra factor N(t1,s) and a bound on N1/N2 that is not established.","rationale":"The reader's weakest assumption was Theorem 1.8; I agree that the rigidity result is load-bearing. But a more concrete, checkable gap is the scaling mismatch in the proof of the Key Matching Proposition, which is downstream of Theorem 1.8 and directly supports Theorem A. This is not an objection to the strategy—the gap is likely repairable by carrying the factor N(t1,s) through Lemma 7.2—but as written the proof of Proposition 5.1 does not cover the range claimed unless additional control on N is supplied. For this reason the verdict stays CONDITIONAL: the main theorems are plausible and the fix appears local, but the current manuscript needs a correction or a precise normalization argument in §7.3.1.","tokens_in":50479,"tokens_out":34268,"duration_ms":373094,"concrete_test":"Re-derive the estimate in §7.3.1 with the correct scaling: let a = r N(t1,s1), set y1 = Trem(x1, a v1), y2 = Trem(x2, a v2), and check whether Lemma 7.2 yields dist(g_T y1, g_T y2) ≤ ε under a < δ. Since y2 = Trem(x2, a v2) still lies in g_{t2}T(ω,β), the conclusion would follow if a < δ and the matching gives dist(x1,x2), ||v1-v2|| < θ. Then identify where the factor N(t1,s1) is lost in the current proof: replace the sentence 'y_i = Trem(x_i, r v_i)' by the correct expression and see whether the range r < δ/N(t1,s) or a uniform lower bound on N(t1,s) over S_ℓ is needed. If the argument requires a lower bound that fails for KZ-contracted directions, Proposition 5.1 is unproved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Proposition 5.1 is the key input for Theorem A. In §7.3.1 the proof sets v_i = β(t_i,s_i)/N(t_i,s_i) and y_i = Trem(x_i, r v_i), then applies Lemma 7.2 to conclude dist(g_T y1, g_T y2) ≤ ε for r < δ/N(t1,s). But by the standing notation (2.11), Tremβ(t,s,r) = Trem(ω(t,s), r·β(t,s)), so the actual tremor is Trem(x1, r N(t1,s1) v1), not Trem(x1, r v1). The stated range r < δ/N is designed to make the actual displacement rN(t1,s1) < δ. To use Lemma 7.2 one must compare Trem(x1, rN1 v1) with Trem(x2, rN1 v2) for the same displacement rN1; however the proof compares Trem(x1, r v1) with Trem(x2, r v2), which are different points in the torus and have displacement r, not rN1. The missing factor N(t1,s1) also introduces a term r|N1-N2| in period coordinates, and the argument gives no control of |N1-N2| on the matching set. In particular, if N(t1,s) is not uniformly bounded below on K (and the KZ cocycle can contract directions in Cyl0, so this is not automatic), the proof does not establish item (2) for the full stated range.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the distribution of expanding twist tori g_t·T(ω) in moduli spaces of translation surfaces, where T(ω) is the torus of independent horizontal cylinder twists of a horizontally periodic translation surface. The main results are Theorem A, asserting that if (M,ω) is a Veech surface and is M-primitive for M = SL2(R)·T(ω), then g_t·T(ω) becomes dense in M as t→∞ without passing to subsequences, and Theorem B, asserting that under a transverse monodromy hypothesis (1.2) every weak-* limit of the uniform measures (g_t)_*μ_T is fully supported in M. The proofs are built on a strengthening of Forni's full-density equidistribution result (Theorem 1.6), a measure rigidity theorem for U-invariant measures on projective Kontsevich-Zorich bundles (Theorem 1.8), a Key Matching Proposition 5.1, and, for Theorem B, a matching proposition 8.2 and a measure-theoretic matching proposition 8.4. Section 3 supplies infinite families of examples: regular 2n-gon Veech surfaces for Theorem A and the Matheus-Yoccoz square-tiled surfaces for Theorem B.","tokens_in":50767,"tokens_out":14696,"duration_ms":162766,"significance":"If the results are correct, they provide the first no-subsequence statements for expanding twist tori in the non-square-tiled setting and, in Theorem B, a full-support statement for all weak-* limits that is stronger than mere density and new even in previously studied square-tiled cases. The architecture of the proof is coherent and the paper explicitly separates the algebraic hypotheses (M-primitivity and the monodromy hypothesis (1.2)) from the dynamical matching arguments. The examples in Section 3 are concrete and non-trivial, and the paper is unusually candid about which parts are sketches or outlines. That said, two load-bearing pieces are not fully established as written: the proof of the Key Matching Proposition has a missing normalization factor in its application of Lemma 7.2, and the proof of Theorem 1.8 is a compressed summary rather than a complete argument. These are repairable, but they prevent the paper from being accepted in its current form.","major_comments":[{"comment":"The application of Lemma 7.2 is inconsistent with the definition of Tremβ in (2.11). The actual tremor is Tremβ(t1,s,r) = Trem(ω(t1,s), r·β(t1,s)) = Trem(x1, rN(t1,s)v1), where v1 = β(t1,s)/N(t1,s). The proof instead defines y1 = Trem(x1, rv1) and y2 = Trem(x2, rv2), so Lemma 7.2 is applied to tremors with displacement r rather than the displacement rN(t1,s) that actually occurs. Moreover, the hypothesis of Lemma 7.2 requires the displacement parameter to be <δ, but r can exceed δ when N(t1,s)<1, and no lower bound on N(t1,s) on the matching set is established. The gap is repairable: set ρ = rN(t1,s), define y1 = Trem(x1, ρv1) and y2 = Trem(x2, ρv2), and then ρ<δ follows from the stated range r<δ/N(t1,s). With this change y2 still lies in g_{t2}·T(ω,β), and Lemma 7.2 gives the desired estimate. But as written, the proof compares the wrong points, and item (2) of Proposition 5.1 is not established. Since Proposition 5.1 is the key input to Theorem A, this must be corrected.","section":"Section 7.3.1, proof of Proposition 5.1"},{"comment":"The proof of Proposition 6.2 is a compressed summary of Ratner's shearing argument. In particular, the step following (6.3), where the authors say that 'Following Ratner, with the aid of Birkhoff's ergodic theorem, we can find s∈I so that the two points ... also belong to the compact set L̂', is the decisive point of the proof but is not justified in the manuscript. The text does not show how Lemma 6.3 is combined with the decomposition u(s)p = u_-(r_p)g_{t_p}u(s_p) to produce a common s of positive measure for which both relevant points lie in L̂, nor does it define the ambient metric on PbV used in (6.2). Because Lemma 7.3, and hence Proposition 7.1 and Proposition 5.1, depend on Theorem 1.8, this argument needs to be written out in full rather than summarized.","section":"Section 6, proof of Proposition 6.2 and Theorem 1.8"}],"minor_comments":[{"comment":"The proof of Lemma 8.11 is omitted with the phrase 'proof is left to the reader'. The lemma is used in the proof of Proposition 8.9, which is needed for Theorem B. A short proof should be included, even though the statement is elementary.","section":"Section 8.4, Lemma 8.11"},{"comment":"Proposition B.2 is presented as a sketch and relies on entropy arguments from [BAM+18, Led84]. This may be acceptable because it is used only for Remark 1.9 and Theorem B.1 rather than for Theorems A and B, but the paper should state explicitly that this is a sketch and identify the precise statements it invokes.","section":"Appendix B, Proposition B.2"},{"comment":"Theorem A.1 for the decagon is advertised as an example, but its proof is an outline: the substitute sets R^D_{δ0,η} are introduced and Lemma A.6 is stated, but several steps are summarized rather than fully proved. If this appendix is meant to be a full proof, it needs expansion; otherwise it should be labelled as a sketch.","section":"Appendix A, Theorem A.1"},{"comment":"The metric dist on subsets of PbV used in (6.2) is not defined in Section 6; a fiber metric is defined in (6.1), but a global product metric on the bundle is only introduced later, in Section 7.2. Please define the metric used in the measure rigidity argument.","section":"Section 6, equation (6.2)"},{"comment":"The notation Trem(t,s,r) without the subscript β is used after Remark 2.12, and in Section 7.3.1 the same symbol r is used both for the coefficient in Tremβ and for the displacement parameter in Lemma 7.2. This overloading is a source of the gap described in the first major comment and should be clarified.","section":"Throughout, cf. Remark 2.12"}],"recommendation":"major_revision","confidential_remarks":"The paper contains strong new ideas and the main theorems are likely correct, but the proof of the Key Matching Proposition has a concrete normalization error that must be fixed, and the proof of Theorem 1.8 is currently too compressed for the role it plays. I do not see circularity: the hypotheses (1.2) and M-primitivity are algebraic and independent of the conclusions, and although some cited background is by the authors themselves, the central claims do not reduce to those citations. I would encourage the editor to request a careful revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know two things. The paper is a serious piece of work: it gets no-subsequence density and full-support statements for expanding twist tori in new settings, and it packages a reusable measure-rigidity theorem for cocycle skew products. The other thing is that the stress-test note about Proposition 5.1 is correct: in §7.3.1 the proof writes y_i = Trem(x_i, r v_i) where the actual tremor in the statement is Trem(x_i, r N(t_i,s_i) v_i). The factor N is missing. I do not think this sinks the argument. The intended proof works if you set y_i = Trem(x_i, r N(t1,s) v_i), because the target torus contains tremors in all scalar multiples of the matched direction, and the bound from Lemma 7.2 then applies with displacement rN(t1,s) ≤ δ. The proof should be rewritten with that factor explicit; a referee should ask for it.\n\nWhat is genuinely new: Theorem A gives full-time density of g_t·T(ω) under M-primitivity, with no subsequence, and Theorem B gives uniform lower mass bounds and full support under the monodromy hypothesis (1.2). Theorem 1.6 improves Forni's full-density convergence by making the uniformity hold over compact sets, and Theorem 1.8 is a clean statement that U-invariant measures on projective KZ bundles over Veech curves are A-invariant under slow-growth assumptions. The examples—regular 2n-gons for Theorem A, Matheus–Yoccoz square-tiled surfaces for Theorem B—are real and non-vacuous. The architecture follows EMM15 plus a matching argument, which is the right strategy.\n\nSoft spots, in proportion: the scaling slip in §7.3.1 is the only thing I would call an actual error in the written proof, and it is fixable. Some supporting material is genuinely sketched: Lemma 8.11 is left to the reader, Appendix A is an outline, and Proposition B.2 is a sketch. The core rigidity theorem, Theorem 1.8, relies on a substantial Ratner-style adaptation and is not machine-checked, but the argument is coherent and the subpolynomial divergence lemma is plausible. The citation pattern is fine; the authors cite their own earlier work where it is the natural background, not as a substitute for the main proof.\n\nBottom line: this paper deserves a serious referee. The main claims are important, the proofs are detailed enough to check, and the one real flaw I found is a localized scaling mistake, not a load-bearing contradiction. I would send it out.","headline":"Substantial, plausible progress on density and full support of expanding twist tori, with one real scaling slip in the proof of Proposition 5.1 that looks routine to fix.","tokens_in":51349,"tokens_out":4659,"would_cite":true,"duration_ms":57554,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["37D40","37A17","32G15"],"pacs":[],"model":"deepseek-v4-flash","headline":"Expanding twist tori of Veech surfaces become dense in the full orbit closure as $t\\to\\infty$.","keywords":["Veech surfaces","twist tori","Teichmüller geodesic flow","horocycle flow","measure rigidity","Kontsevich–Zorich cocycle","moduli spaces of abelian differentials","equidistribution"],"falsifier":"Look for a $U$-invariant probability measure on the projective Kontsevich–Zorich bundle $\\mathbb{P}\\widehat{V}$ that projects to the Haar measure $\\mu_V$ on $V$ but is not $A$-invariant, while satisfying the polynomial growth bound (1.4); Theorem 1.8 declares such a measure impossible. Alternatively, for a regular $2n$-gon surface with $n>5$, exhibit a compact set $K\\subset\\mathcal{M}$ and $\\varepsilon>0$ such that $g_t\\cdot\\mathbb{T}(\\omega_n)\\cap K$ fails to be $\\varepsilon$-dense for arbitrarily large $t$, contradicting Theorem A.","tokens_in":50226,"feed_emoji":"🌀","tokens_out":9397,"duration_ms":90007,"temperature":0.7,"pith_summary":"Given a translation surface whose horizontal foliation is periodic, the twist torus $\\mathbb{T}(\\omega)$ collects all surfaces obtained by shearing each horizontal cylinder independently. This paper studies how the tori $g_t\\cdot\\mathbb{T}(\\omega)$ spread out under the Teichmüller geodesic flow, aiming to understand whether they equidistribute in the conjectured locus $\\mathcal{M}=\\overline{\\mathrm{SL}_2(\\mathbb{R})\\cdot\\mathbb{T}(\\omega)}$. The main theorems give sufficient conditions for these expanding tori to be dense in $\\mathcal{M}$ for all sufficiently large $t$ (Theorem A), and for every weak-* limit of their uniform measures to have full support in $\\mathcal{M}$ (Theorem B). Because earlier results only worked along subsequences, these are the first no-subsequence statements for non-square-tiled Veech surfaces, and they make the twist-torus analogue of a long-standing conjecture for hyperbolic surfaces plausible in these cases.","feed_headline":"Expanding twist tori become dense in moduli space at all large times","feed_subtitle":"For primitive Veech surfaces, no subsequence passage needed—a step toward the full equidistribution conjecture.","key_machinery":"The central object is the twist torus $\\mathbb{T}(\\omega)$, the compact torus $\\mathbb{R}^n/\\prod m_i\\mathbb{Z}$ of surfaces obtained by applying the horocycle flow to each horizontal cylinder independently. The load-bearing mechanism is a key matching proposition (Proposition 5.1) that compares, near the Veech curve $V=\\mathrm{SL}_2(\\mathbb{R})\\cdot\\omega$, a small piece of $g_t\\cdot\\mathbb{T}(\\omega)$ with the $g_{T+\\ell}$-image of a nearby tremor; the comparison is made via a linear A-invariance statement (Theorem 1.8) for the projective bundle of the Kontsevich–Zorich cocycle over $V$. This A-invariance is proved by adapting the classical shearing argument for unipotent flows, using subpolynomial fiber divergence: under the polynomial growth bound (1.4), the dominant direction of divergence of two nearby points under the horocycle flow is parallel to the base, which forces every $U$-invariant measure projecting to Haar on $V$ to be $A$-invariant. For Theorem B, the matching is promoted from closeness to exact equality of cocycle matrices (Proposition 8.2), which requires the monodromy hypothesis (1.2).","core_discovery":"On the paper's own terms, the discovery is that dense-in-time behavior of expanding twist tori can be pushed through from the (well-understood) horocycle dynamics on the Veech curve to the full orbit closure, using a rigidity theorem for the Kontsevich–Zorich cocycle. Theorem A asserts that if $(M,\\omega)$ is horizontally periodic Veech and $M$-primitive in $\\mathcal{M}$, then for every compact $K\\subset\\mathcal{M}$ and $\\varepsilon>0$, the set $K\\cap g_t\\cdot\\mathbb{T}(\\omega)$ is $\\varepsilon$-dense in $K$ for all $t$ large enough. Theorem B asserts that under hypothesis (1.2)—a pseudo-Anosov element of the affine group acting as the identity on $\\mathrm{Cyl}^0(\\omega)$—every weak-* limit of $(g_t)_*\\mu_{\\mathbb{T}}$ has full support in $\\mathcal{M}$, with a uniform lower bound on the mass of any nonempty open set. The paper also shows both hypotheses are satisfied by infinite families of classical examples, and that the decagon, though not primitive, satisfies the density conclusion by an appendix argument.","pith_inferences":["The paper's Remark 1.9 and Appendix B suggest that under extra hypotheses on the cocycle (bounded image, or proximal and irreducible), the limiting measure is unique; one could test whether the full convergence in Conjecture 1.5 holds for the exceptional square-tiled family, where those extra hypotheses are naturally satisfied.","A concrete numerical check on the decagon or a regular $14$-gon surface—computing $\\varepsilon$-density of $g_t\\mathbb{T}(\\omega)$ in a fixed ball for a mesh of times—would provide independent confirmation of the density claim in a case not covered by the general primitivity hypothesis.","The uniform lower mass bound in Theorem B resembles an effective equidistribution statement; if made quantitative, it may yield error estimates for counting problems of saddle connections and periodic orbits, mirroring how effective unipotent flow results feed counting in flat geometry."],"forward_implications":["Under Theorem A, the torus $g_t\\cdot\\mathbb{T}(\\omega)$ must intersect every open subset of $\\mathcal{M}$ for every sufficiently large $t$, eliminating the possibility of escape to proper invariant submanifolds in the primitive case.","Under Theorem B, the uniform measures on $g_t\\cdot\\mathbb{T}(\\omega)$ cannot concentrate on any proper closed subset: each nonempty open set receives mass at least some uniform $\\varepsilon>0$ for all large $t$.","The hypotheses of both theorems are checkable in concrete families, including the regular $2n$-gon surfaces for $n>5$ and the exceptional square-tiled family of examples, so the density and full-support conclusions apply to infinitely many distinct Veech curves.","The correspondence between flat and hyperbolic twist tori means that full convergence of these flat tori would settle the corresponding conjecture for hyperbolic surfaces along the relevant sequences.","Theorem 1.8 provides a new measure rigidity statement for general locally constant cocycles over $\\mathrm{SL}_2(\\mathbb{R})/\\Gamma$ with polynomial growth, which is of independent use in homogeneous dynamics."],"supporting_citations":[{"why":"Supplies the orbit-closure and isolation results underlying Theorem 1.6 and the support of $\\mu_{\\mathcal{M}}$.","marker":"[EMM15]"},{"why":"Establishes full-density convergence of $g_t$-pushes of horocycle arcs, the prior result Theorem 1.6 strengthens.","marker":"[For21]"},{"why":"Provides the shearing argument adapted in Section 6 to prove A-invariance of $U$-invariant measures on the projective bundle (Theorem 1.8).","marker":"[Rat92]"},{"why":"Provides AGY norm estimates and non-uniform hyperbolicity (Propositions 2.4, 2.7, 2.9) used to control tremor paths and contraction.","marker":"[AG13]"},{"why":"Introduces tremor deformations and balanced cohomology classes, the language in which twist tori and $\\mathrm{Cyl}^0(\\omega)$ are defined.","marker":"[CSW20]"},{"why":"Gives equivariance of horizontal cylinder twists under $g_t$ (Lemma 2.13) and distance-to-norm estimates used in matching.","marker":"[CKS21]"},{"why":"Constructs the square-tiled family whose affine group has a pseudo-Anosov element acting trivially on the complement of the tautological plane, the examples for Theorem B.","marker":"[MY10]"},{"why":"Shows finiteness of Veech curves with trace field of degree at least 3, used to prove the twist torus of regular $2n$-gons meets finitely many closed $\\mathrm{SL}_2(\\mathbb{R})$-orbits.","marker":"[EFW18]"},{"why":"Provides classification of orbit closures in genus two, used in the decagon appendix and in establishing primitivity of the $2n$-gon examples.","marker":"[McM07]"}],"fun_headline_variants":["Twist tori go dense in moduli space, no subsequence trick","Dense twist tori for all large times on Veech surfaces","Expanding twist tori: full density without subsequences","Veech surfaces: twist tori hit every open set eventually"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The central load-bearing premise is the rigidity theorem (Theorem 1.8): every $U$-invariant probability measure on the projective Kontsevich–Zorich bundle over a Veech curve that projects to Haar measure on the curve is also invariant under the geodesic flow; if this classification fails, the key matching proposition and with it Theorem A collapse.","fun_headline_variants_meta":{"raw":{"variants":["Twist tori go dense in moduli space, no subsequence trick","Dense twist tori for all large times on Veech surfaces","Expanding twist tori: full density without subsequences","Veech surfaces: twist tori hit every open set eventually"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000435,"raw_usage":{"total_tokens":2286,"prompt_tokens":1086,"completion_tokens":1200,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":702,"completion_tokens_details":{"reasoning_tokens":1124}},"tokens_in":702,"tokens_out":1200,"duration_ms":10936,"temperature":1.0,"reasoning_tokens":1124,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T17:48:33.650305+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Look for a $U$-invariant probability measure on the projective Kontsevich–Zorich bundle $\\mathbb{P}\\widehat{V}$ that projects to the Haar measure $\\mu_V$ on $V$ but is not $A$-invariant, while satisfying the polynomial growth bound (1.4); Theorem 1.8 declares such a measure impossible. Alternatively, for a regular $2n$-gon surface with $n>5$, exhibit a compact set $K\\subset\\mathcal{M}$ and $\\varepsilon>0$ such that $g_t\\cdot\\mathbb{T}(\\omega_n)\\cap K$ fails to be $\\varepsilon$-dense for arbitrarily large $t$, contradicting Theorem A.","supporting_citations":[],"review_version":1}