{"id":"69214c72-88e4-45a0-9d10-5875ae71fa13","arxiv_id":"1908.03253","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Every finite type curve in R3 is a parabolic-point-free asymptotic line of a suitably constructed plane field, with an explicit hyperbolic closed example given by (sin x, cos x, sin 3x).","lead":"Any curve of finite type can be made into an asymptotic line of some plane field in three-dimensional space, and the paper gives a concrete closed hyperbolic example. The result extends Arnold's theorems from surfaces to plane fields, where the extra freedom makes the construction possible.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 3.1's proof does not justify passing from the vanishing normal field (2.5) to a regular plane field at finite-type points with m>2.","rationale":"The reader's weakest assumption identifies exactly the load-bearing gap: for a finite type point with m>2, the constructed normal field vanishes at the point, and the proof's factoring of x^{m−2} from the restricted coefficients b(x,0,0) and c(x,0,0) does not establish that the full plane field is regular in a neighborhood. This is a genuine correctness gap in the main theorem, not merely a matter of exposition. However, it looks fixable: because γ''(0)=0, the osculating plane has a well-defined limit at the origin, so a regular plane field extending the osculating-plane distribution should exist; one only needs to specify the tail functions and second-order coefficients in (2.5) with the right vanishing orders. Thus the correct verdict is CONDITIONAL, matching the reader. The explicit closed example in Theorem 4.3 has nonzero curvature along the curve, so it is not vulnerable to this particular gap. No independent verification or machine-checked proof is present; the explicit formulas are at least checkable by direct symbolic computation.","tokens_in":10429,"tokens_out":11785,"duration_ms":121182,"concrete_test":"Test the missing step on the minimal example γ(x)=(x,x^3,x^4) (m=3,n=4). Instantiate the full ansatz (2.5) with arbitrary smooth free functions, enforce that the quotient ξ/x extends C^1 to x=0 and is nonzero at 0, and recompute a,b,c in (2.3) and K=eg−f^2 on γ. The theorem's proof is complete only if there exists a choice of the free functions with ξ/x(0)≠0, c(0,0,0)≠0, and K(x,0,0)=−1; a symbolic computation would settle whether such a choice exists. Equivalently, check the divisibility identity a(x,y,z)=x \\tilde a(x,y,z) for the full coefficient a, since without divisibility of a the divided equation used in the proof is undefined.","verdict_should_be":"UNCHANGED","load_bearing_attack":"For a finite type curve with symbol (1,m,n), m>2, the vector field ξ defined by (2.5)-(2.6) restricts on γ to ξ(x,0,0)=l0(x)Y(x)+k0(x)Z(x), and (2.6) gives k0(x)=a_m m(m−1)x^{m−2}+O(x^{m−1}), l0(x)=O(x^{n−2}). Hence ξ(0)=0. Since rescaling by a nonvanishing function cannot make a zero vector nonzero, the proof must replace ξ by a different representative of the plane field on the punctured neighborhood, e.g. divide by x^{m−2}. The proof only states that the restricted coefficients b(x,0,0) and c(x,0,0) factor as x^{m−2}B(x,0,0), x^{m−2}C(x,0,0) with C(0,0,0)=a_m m(m−1); it does not show that the full coefficients a,b,c in (2.3), or the full vector field (2.5), are divisible by x^{m−2}. The sentence 'after factoring x^{m−2} from the first equation of (2.3)' is therefore unjustified: an equation a dx + b dy + c dz = 0 cannot be divided by a factor present in only two of the three coefficients unless a is also divisible. No choice of the free tail functions A,B,C and second-order coefficients in (2.5) is specified that would make the divided vector field smooth and nonvanishing at the origin. Without this, the finite type point is not shown to lie on a genuine (regular) plane field, and the subsequent application of Proposition 2.11 and the computation K(x,0,0)=−1 are not established at that point. This is load-bearing for Theorem 3.1 but does not affect the m=2 case or the explicit curve in Theorem 4.3, where γ''(0)≠0.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies ξ-asymptotic lines of plane fields in R^3, i.e., curves whose tangent direction has zero normal curvature with respect to a plane field. The main result, Theorem 3.1, claims that every finite type curve is a ξ-asymptotic line without parabolic points of a suitable plane field. The construction uses a tubular neighborhood of the curve and an explicit vector field ξ given by (2.5), with coefficients chosen so that the Gaussian curvature K = eg - f^2 satisfies K(x,0,0) = -1. The paper also gives Theorem 4.3, an explicit closed hyperbolic finite type example γ(x) = (sin x, cos x, sin 3x), with a computation of the derivative of the Poincaré map and its eigenvalues. An appendix gives a new proof of Arnold's theorem on finite type asymptotic lines of hyperbolic surfaces.","tokens_in":69,"tokens_out":6309,"duration_ms":121378,"significance":"If the main theorem is correct, it generalizes Arnold's classical results by showing that every finite type curve, including curves with inflections, can be realized as an asymptotic line (without parabolic points) of some plane field; the explicit closed hyperbolic example in Theorem 4.3 is a concrete, checkable contribution. The paper's construction is explicit and the computations for the example are detailed. However, the proof of Theorem 3.1 contains a load-bearing gap for finite type points with symbol (1,m,n), m>2, where the constructed vector field vanishes; without a repair of that gap, the main theorem is not established in full generality.","major_comments":[{"comment":"The step \"after factoring x^{m-2} from the first equation of (2.3)\" is not justified. For a finite type point with symbol (1,m,n) and m>2, equations (2.5)-(2.6) give ξ(x,0,0) = l0(x)Y(x) + k0(x)Z(x), with k0(x) = a_m m(m-1)x^{m-2} + O(x^{m-1}) and l0(x) = O(x^{n-2}), so ξ(0,0,0) = 0. The proof shows only that the restricted coefficients b(x,0,0) and c(x,0,0) factor by x^{m-2}; it does not show that the coefficient a(x,y,z) in the 1-form a dx + b dy + c dz is divisible by x^{m-2} in a neighborhood, nor that the vector field (2.5) can be rescaled by a globally defined nonvanishing function to remove the zero. Dividing the 1-form by a factor present in only two of its three coefficients requires divisibility of a as a smooth function; otherwise the operation is not defined. Without this, Proposition 2.11 and the conclusion K(x,0,0) = -1 are not established at the finite type point, so the central claim of Theorem 3.1 is not proved for m>2.","section":"Section 3, proof of Theorem 3.1"},{"comment":"The definition of k1 immediately before the final sentence of the proof divides by the factor ((γ'_1)^2 + (γ'_2)^2 + (γ'_3)^2)(γ'_1 γ''_2 - γ'_2 γ''_1). At a finite type point with symbol (1,m,n) and m>2, the second factor satisfies γ'_1(0) γ''_2(0) - γ'_2(0) γ''_1(0) = 0, so the displayed expression is singular at x=0. The proof does not specify a choice of the free function l1 that makes k1 smooth, nor does it prove that such a choice exists. Thus the constructed plane field is not shown to be smooth (or even continuous) in a neighborhood of the origin in the m>2 case. This is a second load-bearing gap in the same theorem.","section":"Section 3, formula for k1 in Theorem 3.1"}],"minor_comments":[{"comment":"The displayed formula for k1 is typographically ambiguous: the term containing l1 does not clearly indicate its denominator. Since the proof depends on this formula, the authors should rewrite it with unambiguous parentheses and fractions.","section":"Theorem 3.1 and Proposition 4.1"},{"comment":"The notation [β_u, β_v, β_uu] is used without definition; it should be defined as the mixed product (determinant) of the three vectors.","section":"Appendix A"},{"comment":"The definition of a finite type point requires 1 < m < n, but the subsequent sentence says that if n = m+1 then γ is of rotating type; this contradicts the strict inequality and should be rephrased, for example by saying that finite type includes the rotating case as a special limit or by separating the definitions.","section":"Definition 2.3"},{"comment":"In the formula for k1 in the proof of Theorem 4.3, the denominator is a function of cos x; the authors should state explicitly that this denominator is nonvanishing for all x, since the formula otherwise could be singular on the closed curve.","section":"Section 4.2"}],"recommendation":"major_revision","confidential_remarks":"The paper's abstract claims a generalization of Arnold's theorems for plane fields, but the precise relation to Arnold's results for surfaces is only sketched. The appendix gives a new proof of Arnold's theorem whose connection to the main plane-field construction is not fully integrated. These are scope and presentation concerns. The main technical issue for the referee is the m>2 gap in Theorem 3.1; it is likely repairable by choosing the free functions A, B, C, and l1 appropriately, but the current manuscript does not provide such a choice or a proof of existence."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The upshot: this is a reasonable, mostly constructive paper that extends Arnold's surface theorem to plane fields, and it contains a genuinely explicit hyperbolic closed example. The main Theorem 3.1 has a real gap for finite-type points with m>2; I think the theorem is probably true and fixable, but as written it is not established.\n\nWhat's new: an existence statement for arbitrary finite type curves as xi-asymptotic lines without parabolic points, and the explicit closed curve gamma(x)=(sin x, cos x, sin 3x) in Theorem 4.3. The machinery in Section 2, including normal curvature, asymptotic directions, the parabolic set, and the Poincare map derivation, is standard but cleanly assembled. Proposition 4.1 gives a useful sufficient condition with K=-(H)^2, and the eigenvalue computations in Theorem 4.3 look self-consistent. The appendix proof of Arnold's theorem is a bonus; I did not audit it carefully.\n\nThe soft spot is exactly what the stress-test flags. In Theorem 3.1, for a curve of symbol (1,m,n) with m>2, the normal vector xi in (2.5) vanishes at the origin: k0 is order m-2 and l0 is higher order. The proof computes b(x,0,0)=x^{m-2}B and c=x^{m-2}C with C(0)!=0, then says that after factoring x^{m-2} from the first equation of (2.3) one gets a regular plane field. But a 1-form a dx + b dy + c dz is divisible by x^{m-2} only if a is also divisible as a smooth function of (x,y,z); a(x,0,0)=0 is not enough, and no choice of tail coefficients is given to ensure it. So the passage from the vanishing xi to a genuine nonvanishing plane field near the origin is unjustified. This is load-bearing for Theorem 3.1. It does not affect the m=2 case, and the explicit curve in Theorem 4.3 has nonvanishing curvature, so that example stands. The global extension of the local plane field to all of R3 is also not discussed, though I expect that is routine.\n\nFor the right reader, someone in qualitative theory of asymptotic lines or plane fields, this is worth engaging with, and it deserves a serious referee. The referee should be asked to close the m>2 gap. Citation pattern looks fine; the result is genuinely new relative to Arnold. I would not cite it in my own work right now, but I would bring it to a reading group.","headline":"A solid local existence construction for plane-field asymptotic lines, with a real gap in Theorem 3.1 for m>2 that is probably fixable; the explicit hyperbolic closed example stands.","tokens_in":11362,"tokens_out":6738,"would_cite":false,"duration_ms":69425,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["53C12","37C27","34C25"],"pacs":[],"model":"deepseek-v4-flash","headline":"Any finite type curve is an asymptotic line of a suitable plane field.","keywords":["finite type curve","ξ-asymptotic line","plane field","parabolic point","normal curvature","Poincaré map","hyperbolic closed line","Arnold theorem"],"falsifier":"Take the concrete finite type curve $\\gamma(x)=(x,x^m,x^n)$ with $m>2$, substitute the general ansatz (2.5), impose $k_0,l_0$ from (2.6), and attempt to solve for a smooth nonzero vector field $\\xi$ near the origin. If for every choice of the tail functions $A,B,C$ the coefficients after division by $x^{m-2}$ still vanish at the origin or fail to be smooth, then Theorem 3.1 fails at finite type points with $m>2$.","tokens_in":10010,"feed_emoji":"🔁","tokens_out":7420,"duration_ms":77984,"temperature":0.7,"pith_summary":"The paper proves that, in $\\mathbb{R}^3$, the class of curves that can appear as asymptotic lines of a plane field is much larger than the class allowed by surfaces: every finite type curve is a $\\xi$-asymptotic line without parabolic points of some plane field. This extends Arnold's theorem for asymptotic lines on hyperbolic surfaces to the setting of arbitrary plane fields, where the field is constructed around the curve rather than prescribed in advance. The proof works in a tubular neighborhood of the curve, builds a normal vector field whose zero curvature condition holds along the curve, and chooses free coefficients so that the discriminant $eg-f^2$ is negative there. A second result gives a concrete closed example, $\\gamma(x)=(\\sin x,\\cos x,\\sin 3x)$, and shows that the associated closed $\\xi$-asymptotic line is hyperbolic via the eigenvalues of its Poincaré return map.","feed_headline":"Any finite type curve is an asymptotic line of a suitable plane field","feed_subtitle":"Generalizes Arnold's surface theorem and gives a hyperbolic closed example on a trefoil-like curve.","key_machinery":"The load-bearing object is the tubular neighborhood framing $\\alpha(x,y,z)=\\gamma(x)+yY(x)+zZ(x)$ with $Y=(\\gamma_2',-\\gamma_1',0)$ and $Z=X\\wedge Y$, together with the general ansatz (2.5) for a vector field $\\xi$ that makes $\\gamma$ an integral curve of the plane field. In these coordinates the $\\xi$-asymptotic line condition becomes the implicit system $a\\,dx+b\\,dy+c\\,dz=0$ and $L_1dx^2+\\cdots+L_6dz^2=0$; when $c\\neq 0$ this reduces to $dz=-(a/c)dx-(b/c)dy$ and the quadratic equation $e\\,dx^2+2f\\,dx\\,dy+g\\,dy^2=0$, whose discriminant $K=eg-f^2$ is the plane-field analogue of Gaussian curvature. The proof's engine is the factorization $b=x^{m-2}B$, $c=x^{m-2}C$ with $C(0,0,0)\\neq 0$, which removes an apparent singularity at the finite type point, and the choice of the free coefficient $k_1$ forcing $K(x,0,0)=-1$. For the closed example, the same machinery feeds the linearization of the Poincaré map, $dP(0,0)=Q(l)$ from the matrix system (4.3).","core_discovery":"The central claim is Theorem 3.1: any finite type curve $\\gamma(x)=(x,a_m x^m+O_{m+1}(x),a_n x^n+O_{n+1}(x))$, with $1<m<n$ and $a_m a_n\\neq 0$, is a $\\xi$-asymptotic line without parabolic points of a suitable plane field. The proof places the curve in the tubular coordinates $\\alpha(x,y,z)=\\gamma(x)+yY(x)+zZ(x)$, where $Y=(\\gamma_2',-\\gamma_1',0)$ and $Z=X\\wedge Y$, and writes a general vector field $\\xi$ in the form (2.5). The condition that $\\gamma$ is an asymptotic line fixes the low-order coefficients $k_0,l_0$ in terms of $\\gamma'$ and $\\gamma''$. The decisive algebraic step is that the coefficients $b(x,0,0)$ and $c(x,0,0)$ of the implicit equation share the factor $x^{m-2}$; after dividing it out, $c(0,0,0)=a_m m(m-1)\\neq 0$, so the equation can be solved for $dz$, and the remaining freedom in the vector field is used to make the plane-field curvature $K=eg-f^2$ equal to $-1$ along the curve. The paper then computes, for the closed curve $\\gamma(x)=(\\sin x,\\cos x,\\sin 3x)$, the eigenvalues of $dP(0,0)$ as $e^{2\\pi}$ and $e^{-25\\pi/8}$, neither lying on the unit circle, proving that the closed $\\xi$-asymptotic line is hyperbolic.","pith_inferences":["The same tubular-neighborhood ansatz can likely prescribe not only the curve but also the value of the plane-field curvature along it, for instance $K=-H^2$ from Proposition 4.1, suggesting that the parabolic set of the field can be engineered to lie away from a prescribed finite type curve.","The factorization step indicates that finite type points with $m>2$ are not singular for plane fields in the way they are for surfaces; a natural test is whether the construction extends to symbols with $m=2$ or to rotating type curves $n=m+1$ by the same division argument.","If the theorem is true, the solution set of the implicit equation for $\\xi$-asymptotic lines can be made to contain any prescribed finite type curve as a hyperbolic branch, so plane-field asymptotic foliations are substantially more flexible than surface asymptotic foliations.","A testable extension would be the higher-dimensional analogue: whether every finite type curve in $\\mathbb{R}^n$ can be realized as an asymptotic line of a suitable hyperplane field in $\\mathbb{R}^n$."],"forward_implications":["Any finite type curve in $\\mathbb{R}^3$, including curves with inflection points, can be realized as a $\\xi$-asymptotic line of a plane field with no parabolic points on the curve, so the curve itself imposes no obstruction once the plane field is allowed to be non-integrable.","Arnold's theorem for asymptotic lines on hyperbolic surfaces becomes the integrable special case of this construction; the appendix gives a new proof of that theorem through the same normal-form calculation.","The explicit curve $\\gamma(x)=(\\sin x,\\cos x,\\sin 3x)$ is a closed hyperbolic $\\xi$-asymptotic line: the eigenvalues of its Poincaré return derivative are $e^{2\\pi}$ and $e^{-25\\pi/8}$, so the closed line is hyperbolic in the sense of periodic orbits.","Because plane fields need not be integrable, closed $\\xi$-asymptotic lines can have convex or starlike projections, which is impossible for closed asymptotic lines on surfaces $z=\\varphi(x,y)$ by the Panov theorem quoted as Theorem 2.2; the circle example in Section 2 already exhibits this flexibility."],"supporting_citations":[{"why":"Supplies the definition of finite type curve and Arnold's theorem that this paper generalizes to plane fields.","marker":"[4]"},{"why":"Provides the normal curvature of a plane field, the implicit equation for $\\xi$-asymptotic lines, and the Jacobi integrability criterion used throughout.","marker":"[2]"},{"why":"Defines hyperbolic closed orbits and Poincaré maps, used to call the closed $\\xi$-asymptotic line hyperbolic.","marker":"[12]"},{"why":"Gives the classical Euler normal curvature concept that the plane-field definition extends.","marker":"[5]"},{"why":"Introduces plane fields in $\\mathbb{R}^3$ as kernels of one-forms, the setting of the paper.","marker":"[14]"}],"fun_headline_variants":["Every finite type curve is an asymptotic line of some plane field","Arnold's theorem extended to plane fields for finite type curves","Hyperbolic closed trefoil: finite type asymptotic line in plane field","Plane fields: every finite type curve has an asymptotic line","Finite type curves become asymptotic lines in suitable plane fields"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"In the proof of Theorem 3.1 the argument divides the common factor $x^{m-2}$ out of the coefficients of the implicit equation, and everything rests on the resulting quotient defining a smooth nonzero plane field in a neighborhood of the finite type point; if no choice of the tail functions makes that quotient smooth and nonvanishing, the construction does not produce a genuine plane field there.","fun_headline_variants_meta":{"raw":{"variants":["Every finite type curve is an asymptotic line of some plane field","Arnold's theorem extended to plane fields for finite type curves","Hyperbolic closed trefoil: finite type asymptotic line in plane field","Plane fields: every finite type curve has an asymptotic line","Finite type curves become asymptotic lines in suitable plane fields"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001151,"raw_usage":{"total_tokens":4773,"prompt_tokens":949,"completion_tokens":3824,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":565,"completion_tokens_details":{"reasoning_tokens":3737}},"tokens_in":565,"tokens_out":3824,"duration_ms":27292,"temperature":1.0,"reasoning_tokens":3737,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:20:55.477152+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the concrete finite type curve $\\gamma(x)=(x,x^m,x^n)$ with $m>2$, substitute the general ansatz (2.5), impose $k_0,l_0$ from (2.6), and attempt to solve for a smooth nonzero vector field $\\xi$ near the origin. If for every choice of the tail functions $A,B,C$ the coefficients after division by $x^{m-2}$ still vanish at the origin or fail to be smooth, then Theorem 3.1 fails at finite type points with $m>2$.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the definition of finite type curve and Arnold's theorem that this paper generalizes to plane fields."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the normal curvature of a plane field, the implicit equation for $\\xi$-asymptotic lines, and the Jacobi integrability criterion used throughout."},{"cited_title":"Palis, Jr","cited_arxiv_id":null,"evidence_quote":"Defines hyperbolic closed orbits and Poincaré maps, used to call the closed $\\xi$-asymptotic line hyperbolic."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the classical Euler normal curvature concept that the plane-field definition extends."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces plane fields in $\\mathbb{R}^3$ as kernels of one-forms, the setting of the paper."}],"review_version":1}