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

The symmetric Mahler conjecture and its equality cases

For every integer n≥1n \ge 1 and every origin-symmetric convex body K⊂RnK \subset \mathbb{R}^n, ∣K∣∣K∘∣≥4n/n!|K||K^\circ| \ge 4^n/n!, with equality if and only if KK is a linear Hanner body.

For every integer n≥1n \ge 1 and every even proper lower-semicontinuous convex function φ:Rn→R∪{+∞}\varphi : \mathbb{R}^n \to \mathbb{R} \cup \{+\infty\} such that 0<∫Rne−φ(x) dx<∞0 < \int_{\mathbb{R}^n} e^{-\varphi(x)}\,dx < \infty, (∫Rne−φ(x) dx)(∫Rne−φ∗(y) dy)≥4n(\int_{\mathbb{R}^n} e^{-\varphi(x)}\,dx)(\int_{\mathbb{R}^n} e^{-\varphi^*(y)}\,dy) \ge 4^n.

For every integer n≥1n \ge 1, if ηi(dx)=e−Vi(x) dx\eta_i(dx) = e^{-V_i(x)}\,dx (i=1,2i=1,2) are full-dimensional origin-symmetric log-concave probability measures with proper lower-semicontinuous convex potentials ViV_i and essentially continuous densities, then their moment measures νi=(∇Vi)#ηi\nu_i = (\nabla V_i)_\# \eta_i have finite first moments, H(ηi)=−∫Vi dηiH(\eta_i) = -\int V_i\,d\eta_i is finite, and H(η1)+H(η2)≤−nlog⁡(4e2)+T(ν1,ν2)H(\eta_1) + H(\eta_2) \le -n \log(4e^2) + \mathcal{T}(\nu_1, \nu_2).

For nonzero, pairwise nonproportional real rows b1,…,bmb_1, \dots, b_m spanning (Rd)∗(\mathbb{R}^d)^*, the strip body AA and feasible-simplex union ΣX\Sigma_X satisfy 1≤S=d!4d∫A∣ΣX∣ dX≤d!4d∣A∣∣A∘∣1 \le S = \frac{d!}{4^d} \int_A |\Sigma_X|\,dX \le \frac{d!}{4^d}|A||A^\circ|, and in particular ∣A∣∣A∘∣≥4d/d!|A||A^\circ| \ge 4^d/d! and 0≤∫A(∣A∘∣−∣ΣX∣) dX≤∣A∣∣A∘∣−4d/d!0 \le \int_A (|A^\circ| - |\Sigma_X|)\,dX \le |A||A^\circ| - 4^d/d!.

If an origin-symmetric convex body K⊂RnK \subset \mathbb{R}^n, n≥1n \ge 1, satisfies P(K)=4n/n!P(K) = 4^n/n!, then KK is a linear Hanner body.

Every dd-dimensional linear Hanner body HH satisfies P(H)=4d/d!P(H) = 4^d/d!, and the polar of a linear Hanner body is again a linear Hanner body.

The power series F(z)=8π2∑j=0∞(−1)jz2j+1(2j+1)2F(z) = \frac{8}{\pi^2} \sum_{j=0}^\infty \frac{(-1)^j z^{2j+1}}{(2j+1)^2} is a biholomorphism from D\mathbb{D} onto a bounded convex domain D={v+it:∣t∣<1,∣v∣<λ(t)}D = \{v + it : |t| < 1, |v| < \lambda(t)\} with F′(0)=8/π2F'(0) = 8/\pi^2, where λ\lambda is even, smooth and strictly concave on (−1,1)(-1,1), and F(eiΘ)=lawελ(T)+iTF(e^{i\Theta}) \stackrel{\mathrm{law}}{=} \varepsilon \lambda(T) + iT for uniform Θ∈[0,2π)\Theta \in [0, 2\pi).

For a holomorphic map f:U→CNf : \mathcal{U} \to \mathbb{C}^N on an open neighborhood U⊂Cd\mathcal{U} \subset \mathbb{C}^d of 00 with f−1(0)={0}f^{-1}(0) = \{0\}, f(z)=H(z)+O(∣z∣k+1)f(z) = H(z) + O(|z|^{k+1}) (HH homogeneous of degree kk, H(z)≠0H(z) \ne 0 for z≠0z \ne 0), and compact sublevels {∣f∣2≤s}\{|f|^2 \le s\} for 0<s<10 < s < 1, the integral of the sum of squared d×dd \times d Jacobian minors over {∣f∣2<1}\{|f|^2 < 1\} is at least (πk)d/d!(\pi k)^d / d!.

For nonzero real rows b1,…,bmb_1, \dots, b_m spanning (Rd)∗(\mathbb{R}^d)^* and any biholomorphism F:D→DF : \mathbb{D} \to D onto a bounded planar domain continuous on D‾\overline{\mathbb{D}} with F(0)=0F(0) = 0, the sum of boundary feasibility probabilities satisfies S=∑I∈BPI≥1S = \sum_{I \in \mathcal{B}} P_I \ge 1.

For an origin-symmetric convex body K⊂RnK \subset \mathbb{R}^n with P(K)=4n/n!P(K) = 4^n/n!, Q=K∘Q = K^\circ, and lifted bodies B=K⊕1[−1,1]B = K \oplus_1 [-1,1], C=Q×[−1,1]C = Q \times [-1,1] in Rn+1\mathbb{R}^{n+1}, for every X∈int⁡BX \in \operatorname{int} B and nonzero ξ=(q,a)∈Rn×R\xi = (q, a) \in \mathbb{R}^n \times \mathbb{R}, there exist 1≤r≤d1 \le r \le d vectors ci=(bi,σi)∈Q×{−1,1}c_i = (b_i, \sigma_i) \in Q \times \{-1, 1\}, nonnegative αi\alpha_i with ∑αi=pC(ξ)\sum \alpha_i = p_C(\xi), and a witness Y∈RdY \in \mathbb{R}^d such that ξ=∑αici\xi = \sum \alpha_i c_i, ∣h⋅Y∣≤λ(h⋅X)|h \cdot Y| \le \lambda(h \cdot X) for all h∈Ch \in C, and ci⋅Y=λ(ci⋅X)c_i \cdot Y = \lambda(c_i \cdot X), minimizing the lens cost ∑αiλ(ci⋅X)\sum \alpha_i \lambda(c_i \cdot X) among all finite positive representations of ξ\xi in CC.

If K⊂RnK \subset \mathbb{R}^n, n≥1n \ge 1, is an origin-symmetric convex body with P(K)=4n/n!P(K) = 4^n/n! and Q=K∘Q = K^\circ, then every triple of points in (Rn,pQ)(\mathbb{R}^n, p_Q) has a metric median.

Methods

Constructing a biholomorphism F(z) = (8/\pi^2) \sum_{j=0}^\infty (-1)^j z2j+1z^{2j+1}/(2j+1)^2 from the unit disk onto a convex planar lens D whose boundary has a uniformly distributed imaginary coordinate under uniform angular measure.

Source excerpts

“Write D = {z ∈ C : |z| < 1} and define F(z) = (8/π^2) ∑_{j=0}^∞ (−1)^j z2j+1z^{2j+1} / (2j + 1)^2.” · Show in paper

Bounding the integral of squared Jacobian minors of a holomorphic map at an isolated zero via Stokes' theorem, Cartan's identity on the unit sphere, and comparison with the homogeneous radial function |z|^{2k}.

Source excerpts

“We first bound the integral of (dd^c τ)^d. The conversion to Jacobian minors will then follow from Cauchy–Binet.” · Show in paper

Applying the holomorphic mass estimate to high powers f_k = (h_1^k, \dots, h_m^k) of the inverse lens map, followed by radial substitution u_i = \rho_i1/ki^{1/k} eiθie^{i\theta_i} and Fatou's lemma, to lower-bound the sum of boundary feasibility probabilities.

Source excerpts

“For an integer k ≥ 1, apply Lemma 4.1 to the holomorphic map fk = (h_1^k, . . . , h_m^k), τk = ∑_{j=1}^m |hj|^{2k}. We shall let k tend to infinity. The mass lower bound has scale k^d. The radial substitution below removes the same factor from the Jacobian integrals and leaves a fixed domain on which the inferred points tend to boundary samples.” · Show in paper

Lifting an equality body K \subset \mathbb{R}^n to B = K \oplus_1 [-1, 1] in \mathbb{R}^{n+1} and using inner polytope approximations with vanishing missing volume on product neighborhoods to extract a common feasibility witness minimizing the lens cost.

Source excerpts

“We now prove Proposition 7.1 in three steps: approximate C by polytopes, locate feasible simplices near each prescribed point and direction, and pass to a common witness to compare costs.” · Show in paper

Comparing the optimal lens-cost representation against a two-term representation as X_\delta = (\delta u, 1-\delta) approaches the apex (0,1) of B, and applying convex separation to obtain metric medians and the 3.2 intersection property for the Hansen–Lima classification.

Source excerpts

“Fix u ∈ int K and let Xδ = (δu, 1 − δ), 0 < δ < 1. ... Apply Proposition 7.1 with this Xδ and (q, a) = (x − y, p1 − p2).” · Show in paper

“Proposition 8.1 and Lemma 8.2 give the 3.2 intersection property for (R^n, pQ). ... Theorem 8.3 gives an invertible linear map T and a Hanner polytope H with Q = T H.” · Show in paper