OctoInformBeta

OpenAI preprint, September 2026 · §3. A regularized matrix field · §3

Lemma 3.3: For σ<1\sigma < 1 and δ>0\delta > 0, explicit evaluation and differentiability on [0,∞)[0,\infty), with strict monotonicity when σ>0\sigma > 0

Formula and shift limit for the scalar comparison function.

Hover over a blue or green term in the text, or a node in the graph, to find it in the other. Select it to read its declaration both in Lean and in words.

The Setting

In the spectral theory of one-dimensional Schrödinger operators with spectral exponent γ∈(1/2,3/2)\gamma \in (1/2, 3/2), the parameter σ=γ−1/2∈(0,1)\sigma = \gamma - 1/2 \in (0, 1) and a regularization parameter δ>0\delta > 0 govern a regularized scalar comparison function built from a normalized integral over ℝ\mathbb{R}. The question is what explicit behavior this comparison function exhibits for nonnegative arguments and under quadratic reflection, and whether shifting the argument inside that integral preserves its value in the symmetric infinite-cutoff limit.

Definitions

For each σ∈ℝ\sigma \in \mathbb{R}, the field normalization constant c(σ)c(\sigma) is the real number defined by c(σ)=σ∫ℝ(1+s2)σ−3/2ds.c(\sigma) = \frac{\sigma}{\int_{\mathbb{R}} (1 + s^2)^{\sigma - 3/2}\,ds}.(1)✦

Given σ,δ∈ℝ\sigma, \delta \in \mathbb{R}, the regularized scalar primitive fσ,δf_{\sigma,\delta} is the function from ℝ\mathbb{R} to ℝ\mathbb{R} given at y∈ℝy \in \mathbb{R} by fσ,δ(y)=∫0y(t++δ)σ−3/2dt,f_{\sigma,\delta}(y) = \int_{0}^{y} (t_+ + \delta)^{\sigma - 3/2}\,dt,(2)✦ where t+=max⁡(t,0)t_+ = \max(t, 0).

The scalar comparison function μσ,δ\mu_{\sigma,\delta} associated with σ,δ∈ℝ\sigma, \delta \in \mathbb{R} is the map from ℝ\mathbb{R} to ℝ\mathbb{R} given at z∈ℝz \in \mathbb{R} by μσ,δ(z)=c(σ)∫ℝ(fσ,δ(s2+z)−fσ,δ(s2))ds,\mu_{\sigma,\delta}(z) = c(\sigma) \int_{\mathbb{R}} \bigl(f_{\sigma,\delta}(s^2 + z) - f_{\sigma,\delta}(s^2)\bigr)\,ds,(3)✦

The Claim

Theorem 1. Let σ,δ∈ℝ\sigma, \delta \in \mathbb{R} satisfy σ<1\sigma < 1 and δ>0\delta > 0. If in addition σ>0\sigma > 0, then for all z1,z2∈ℝz_1, z_2 \in \mathbb{R} with z1<z2z_1 < z_2, μσ,δ(z1)<μσ,δ(z2).\mu_{\sigma,\delta}(z_1) < \mu_{\sigma,\delta}(z_2).(4)✦ At every z∈[0,∞)z \in [0, \infty), μσ,δ\mu_{\sigma,\delta} is differentiable at zz with μσ,δ′(z)=σ(z+δ)σ−1,\mu_{\sigma,\delta}'(z) = \sigma (z + \delta)^{\sigma - 1},(5)✦ and satisfies μσ,δ(z)=(z+δ)σ−δσ.\mu_{\sigma,\delta}(z) = (z + \delta)^\sigma - \delta^\sigma.(6)✦ For all b,z∈ℝb, z \in \mathbb{R}, limR→∞c(σ)∫−RR(fσ,δ((s−b)2+z)−fσ,δ(s2))ds=μσ,δ(z).\lim_{R \to \infty} c(\sigma) \int_{-R}^{R} \bigl(f_{\sigma,\delta}((s - b)^2 + z) - f_{\sigma,\delta}(s^2)\bigr)\,ds = \mu_{\sigma,\delta}(z).(7)✦ For a natural number NN, a vector k=(k1,…,kN)∈ℝNk = (k_1, \dots, k_N) \in \mathbb{R}^N, and an index i∈{1,…,N}i \in \{1, \dots, N\}, x↦μσ,δ(ki2−x2) is locally Lipschitz on ℝ.x \mapsto \mu_{\sigma,\delta}(k_i^2 - x^2) \text{ is locally Lipschitz on } \mathbb{R}.(8)✦

In the paper, the result reads: the function μσ,δ\mu_{\sigma,\delta} is locally Lipschitz and strictly increasing, with μσ,δ(0)=0\mu_{\sigma,\delta}(0) = 0; for z≥0z \ge 0, μσ,δ(z)=(z+δ)σ−δσ\mu_{\sigma,\delta}(z) = (z + \delta)^\sigma - \delta^\sigma and μσ,δ′(z)=σ(z+δ)σ−1\mu_{\sigma,\delta}'(z) = \sigma (z + \delta)^{\sigma - 1}; and for every fixed b,z∈ℝb, z \in \mathbb{R}, c(σ)lim⁡R→∞∫−RR(fσ,δ((s−b)2+z)−fσ,δ(s2))ds=μσ,δ(z)c(\sigma) \lim_{R \to \infty} \int_{-R}^{R} \bigl(f_{\sigma,\delta}((s - b)^2 + z) - f_{\sigma,\delta}(s^2)\bigr)\,ds = \mu_{\sigma,\delta}(z).

By contrast, (8) restricts local Lipschitz continuity to the reflected quadratic coordinate map and (7) retains c(σ)c(\sigma) inside the limit, while σ>0\sigma > 0 is assumed in (4) and not in (5), (6), (7), and (8).

What Is Checked

These claims are stated over the field normalization constant c(σ)c(\sigma) of (1), the regularized scalar primitive fσ,δf_{\sigma,\delta} of (2), and the scalar comparison function μσ,δ\mu_{\sigma,\delta} of (3) for real parameters σ<1\sigma < 1 and δ>0\delta > 0, together with σ>0\sigma > 0 in (4), z≥0z \ge 0 in (5) and (6), b,z∈ℝb, z \in \mathbb{R} in (7), and a vector k∈ℝNk \in \mathbb{R}^N with natural number NN and index i∈{1,…,N}i \in \{1, \dots, N\} in (8).