Singh Theorem:
ProvedSingh.singh_theoremrandom-matrix-theory
For every integer ,
where (cueV N s) is the conditional variance of for two adjacent eigenangles of a Haar unitary at normalized gap . Concretely, with and , is the variance of under the probability density on proportional to .
Preamble
import Mathlib import Definitions.Def_Singh open MeasureTheory Filter Topology
Formal statement
namespace Singh
theorem singh_theorem {N : ℕ} (hN : 2 ≤ N) :
Tendsto (fun s => cueV N s / s ^ 2) (𝓝[>] 0)
(𝓝 (Real.pi ^ 2 / 15 * (1 - 4 / (N : ℝ) ^ 2))) := by sorry
end SinghSource
J. Singh, https://github.com/jjsingh88/singh-constant (commit 792d985)
Read-back
What the Lean code literally says, in plain math · gemini-3.1-pro-preview
This theorem states that for any natural number , the limit as a real variable approaches from the right () of the function is equal to:
The numerator is given by , where for any natural number and real number , the quantity is defined as:
The integrations are taken over with respect to the standard Lebesgue measure. The integrands are defined as follows:
- Difference function: , where . Note that the formal code applies the real logarithm directly to without an absolute value.
- Weight function: , where is the indicator function of the open box (meaning that every coordinate must strictly lie in the interval ).
- Raw weight: is defined as the product:
Degenerate and Edge Cases:
- As is standard in Lean, division by zero evaluates to . This applies to the fractions in if the integral of the weight happens to be zero.
- The real logarithm
Real.log(x)in Lean evaluates to when . This would trigger in if . - The evaluation uses natural number subtraction. Because the theorem explicitly assumes , the term avoids Lean's saturating subtraction (which would cap at ) and behaves identically to standard mathematical subtraction.
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.