By OpenAISourceGitHub

Manuscripts from openai/math

Fork
Clone
git clone https://openversions.com/git/index/openai-math.git
Cite
Version link
https://openversions.com/index/openai-math/commit/dfd8c46870b48511e697cbb43a3c8c91d375fa2f
Version identifier (commit hash)
dfd8c46870b48511e697cbb43a3c8c91d375fa2f

Fork to create your own copy with credit to the source. Clone to get a Git copy on your computer.

All papers

Claims

Findings

Maximal Seshadri constants on arbitrary polarized surfaces

For every smooth integral complex projective surface SS and every ample line bundle LL on SS, there is an integer r0=r0(S,L)r_0 = r_0(S, L) such that for every integer r≥r0r \ge r_0, outside a countable union ZrZ_r of proper Zariski-closed subsets of UrU_r (with nonempty complement), the class π∗L−L2/r∑i=1rEi\pi^* L - \sqrt{L^2/r} \sum_{i=1}^r E_i is nef on the blow-up at p∈Ur∖Zr\mathbf{p} \in U_r \setminus Z_r, and consequently ε(S,L;p)=L2/r\varepsilon(S, L; \mathbf{p}) = \sqrt{L^2/r}.

For any uncountable algebraically closed field KK of characteristic zero, smooth integral projective surface S/KS/K, and ample line bundle LL on SS, there is an integer r0=r0(S,L)r_0 = r_0(S, L) such that for every r≥r0r \ge r_0, outside a countable union of proper Zariski-closed subsets of UrU_r (with nonempty complement), ε(S,L;p1,…,pr)=L2/r\varepsilon(S, L; p_1, \dots, p_r) = \sqrt{L^2/r}.

For a finite-dimensional complex vector space F⊂C{x,z}F \subset \mathbb{C}\{x, z\} of convergent holomorphic germs and t>0t > 0, if the initial monomial space span⁡C{XαZβ:(α,β)∈Et(F)}\operatorname{span}_{\mathbb{C}}\{X^\alpha Z^\beta : (\alpha, \beta) \in E_t(F)\} has an injective order-mm jet test at rr distinct points of (C×)2(\mathbb{C}^\times)^2, then representatives of FF on a common neighborhood of the origin have an injective order-mm jet test at rr distinct points arbitrarily close to the origin.

For the fixed nodal divisor V∈∣dL∣V \in |dL|, coordinates, and frame, every integer k≥1k \ge 1, every t>Tt > T, and every nonzero s∈H0(S,kL)s \in H^0(S, kL) satisfy νt(s)∈kQt\nu_t(s) \in kQ_t, and H0(S,kL)H^0(S, kL) has a basis with pairwise distinct initial exponents in kQtkQ_t.

For positive integers k,m,rk, m, r, if M(kPt)\mathcal{M}(kP_t) has injective order-mm jets at some rr distinct points of (C×)2(\mathbb{C}^\times)^2, then H0(S,kL)H^0(S, kL) has injective order-mm jets at some rr distinct points of SS.

For the fixed parameters d,H,Td, H, T of Section 3, there is an integer r0r_0 such that for every integer r≥r0r \ge r_0 one can choose a single t>Tt > T and a subgroup G⊂(C×)2G \subset (\mathbb{C}^\times)^2 of rr distinct points such that for all positive integers k,mk, m satisfying m>kH/rm > k\sqrt{H/r}, the order-mm jet test on M(kPt)\mathcal{M}(kP_t) at GG is injective.

If for all positive integers k,mk, m with m>kH/rm > k\sqrt{H/r}, no nonzero section of L⊗kL^{\otimes k} vanishes to order at least mm at every pip_i, then on the blow-up π:Y→S\pi: Y \to S at p1,…,prp_1, \dots, p_r, the real divisor class D0=π∗L−H/r∑i=1rEiD_0 = \pi^* L - \sqrt{H/r} \sum_{i=1}^r E_i is nef.

Methods

Transferring injective finite jet tests from initial Taylor monomials to convergent holomorphic germs via perturbed weight separation, analytic rescaling, and persistence of full-rank finite jet matrices.

Source excerpts

“Lemma 2.1 (Persistence of a finite jet test). Let m, r, N be positive integers and let q1, . . . , qr be distinct points of an open subset U ⊂ C2. For 1 ≤ i ≤ N and 0 ≤ s < s0, let fi,s be holomorphic on U. Suppose that, as s ↓ 0, every derivative of fi,s of total order less than m at each qj converges to the corresponding derivative of fi,0. If the jet matrix of f1,0, . . . , fN,0 has column rank N, then the jet matrix of f1,s, . . . , fN,s also has column rank N for all sufficiently small positive s.” · Show in paper

“gi,s(X, Z) = s−λi fi(sX, st′ Z). These functions converge locally uniformly on C2 to Xαi Zβi , in the sense that they are defined on any fixed compact set for all sufficiently small s.” · Show in paper

Bounding initial exponents of global sections of H0(S, kL) in a quadrilateral kQt by intersecting with a fixed integral nodal ample divisor V in |dL|.

Source excerpts

“We now bound the initial exponents of sections of an ample line bundle by intersecting their divisors with one fixed nodal curve. Varying the weight in the two branch coordinates produces a family of quadrilaterals.” · Show in paper

“v(1 + 1/t) ≤ ℓx + ℓz = Ip(D, V ) ≤ D · V = kdH.” · Show in paper

Diagonal compression of exponent slices parallel to (1, 1) to reduce jet tests on the quadrilateral M(kQt) to the triangle M(kPt) of area k^2 H / 2.

Source excerpts

“Use the torus coordinates (x, z) = (u, v/u), (u, v) = (x, xz).” · Show in paper

“For a positive parameter s, substitute v = 1 + sy into (3.6) and divide its member indexed by h by sh. The resulting functions are uq(1 + sy)β− q yh −→ uqyh (s −→ 0+).” · Show in paper

Using an r-point finite subgroup G of (C^x)^2, tuned to a primitive lattice direction u, to decompose Laurent polynomials by character projection and bound their multiplicity at the identity via horizontal slices of a triangle.

Source excerpts

“The test points will form a finite subgroup of the complex torus. Averaging over this group separates a Laurent polynomial into components supported on cosets of a sublattice. In the coordinates of that sublattice, each component lies in a triangle for which a one-point multiplicity estimate applies.” · Show in paper

“For a coset η + M, its character projection is fη = 1/r P g∈G χη(g)−1 τgf.” · Show in paper