Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 3.2 — lower bound for private mean estimation under unknown diagonal covariance

Open
UnknownCovDP.General.Supplement.theorem_3_2

by mikedeng1 · Nov 15, 2026 · Mathlib 0df444a (Lean v4.33.1)

differential-privacylower-boundmean-estimationp2o-batch-stoc27sp2o-gran-per-chapterp2o-plan-paperp2o-v1

Let d,n∈Nd,n\in\mathbb Nd,n∈N with n≥1n\ge1n≥1. Under the prior Pdiag(d)\mathcal P_{\mathrm{diag}}(d)Pdiag​(d), the mean is μ∼N(0,d2Id)\mu\sim N(0,d^2I_d)μ∼N(0,d2Id​) and, independently, Σ=Diag(σ12,…,σd2)\Sigma=\mathrm{Diag}(\sigma_1^2,\dots,\sigma_d^2)Σ=Diag(σ12​,…,σd2​) with the σi2\sigma_i^2σi2​ i.i.d. invχ2(1,1)\mathrm{inv}\chi^2(1,1)invχ2(1,1). Given (μ,Σ)(\mu,\Sigma)(μ,Σ), the dataset is X∼N(μ,Σ)⊗nX\sim N(\mu,\Sigma)^{\otimes n}X∼N(μ,Σ)⊗n. For a prior P\mathcal PP, Rε,δ(P,n)\mathfrak R_{\varepsilon,\delta}(\mathcal P,n)Rε,δ​(P,n) is the Bayes risk: the infimum of the α≥0\alpha\ge0α≥0 such that some (ε,δ)(\varepsilon,\delta)(ε,δ)-differentially private algorithm A\mathcal AA has Pr⁡[∥A(X)−μ∥2≤α]≥3/4\Pr[\|\mathcal A(X)-\mu\|_2\le\alpha]\ge 3/4Pr[∥A(X)−μ∥2​≤α]≥3/4.

There is an absolute constant c>0c>0c>0 such that, whenever nlog⁡n≤cdn\log n\le c\sqrt dnlogn≤cd​:

  1. every (1,c/n)(1,c/n)(1,c/n)-differentially private algorithm A\mathcal AA satisfies
Pr⁡(μ,Σ)∼Pdiag, X∼N(μ,Σ)⊗n, coins of A[∥A(X)−μ∥2≥c d3/2n2log⁡d]≥14;\Pr_{(\mu,\Sigma)\sim\mathcal P_{\mathrm{diag}},\ X\sim N(\mu,\Sigma)^{\otimes n},\ \text{coins of }\mathcal A}\Bigl[\|\mathcal A(X)-\mu\|_2\ge \frac{c\,d^{3/2}}{n^2\log d}\Bigr]\ge\frac14;(μ,Σ)∼Pdiag​, X∼N(μ,Σ)⊗n, coins of APr​[∥A(X)−μ∥2​≥n2logdcd3/2​]≥41​;
  1. for all ε≤1\varepsilon\le1ε≤1 and δ≤c/n\delta\le c/nδ≤c/n,
Rε,δ(Pdiag,n)≥c d3/2n2log⁡d.\mathfrak R_{\varepsilon,\delta}(\mathcal P_{\mathrm{diag}},n)\ge \frac{c\,d^{3/2}}{n^2\log d}.Rε,δ​(Pdiag​,n)≥n2logdcd3/2​.

When the covariance is diagonal but unknown, private mean estimation in the regime nlog⁡n≤dn\log n\le\sqrt dnlogn≤d​ thus costs error of order d3/2/n2d^{3/2}/n^2d3/2/n2 up to a logarithm. This matches the third term of the known upper bound, so that term is necessary.

Differences from the source:

  • encoding: d, n ∈ ℕ with c/n in the DP parameter → 1 ≤ n; reason: c/n is undefined at n = 0; the base mission's accepted encoding (Theorem 2.1)
  • reading: That is, ℜ_{ε,δ}(P_diag, n) = Ω̃(d^{3/2}/n²) → ENNReal.ofReal (c d^{3/2}/(n² log d)) ≤ bayesRisk (Pdiag d) n ε δ with the theorem's own c; reason: the Ω̃ is pinned to the threshold the proof on p. 17 delivers; this is stronger than Ω̃ and a separate conclusion from the probability clause
  • encoding: c d^{3/2}/(n² log d), undefined at d = 1 where log d = 0 → real division with x/0 = 0, so both clauses are trivial at d ≤ 1; reason: Lean convention; no hypothesis d ≥ 2 is added and no instance with d ≥ 2 changes
  • encoding: a randomized algorithm 𝒜 and probability over the prior, the data and the coins of 𝒜 → a Markov kernel κ and the base experimentProb (iterated lintegral); reason: the base mission's model; same instances
  • encoding: (1, c/n)-DP → PrivRobustStat.IsApproxDP (fun X => κ X) 1 (c / n) with replace-one neighbours; reason: the published definition used by the base mission; adds no condition
Preamble
import Mathlib
import Definitions.Def_UnknownCovDP_General_Supplement_DiagPrior

open MeasureTheory ProbabilityTheory
open scoped ENNReal NNReal
Formal statement
namespace UnknownCovDP.General.Supplement

/-- Theorem 3.2, printed p. 16, PDF p. 17. There is `c > 0` such that for all `d, n` with
`n log n ≤ c√d`: every `(1, c/n)`-DP algorithm has
`Pr[‖𝒜(X) − μ‖₂ ≥ c d^{3/2}/(n² log d)] ≥ 1/4` under `(μ,Σ) ∼ P_diag`, `X ∼ N(μ,Σ)^{⊗n}`; and for all
`ε ≤ 1`, `δ ≤ c/n`, `ℜ_{ε,δ}(P_diag, n) ≥ c d^{3/2}/(n² log d)` (the `Ω̃` of the "That is" clause,
pinned to the same threshold). `n ≥ 1` because `c/n` is undefined at `n = 0`. At `d = 1`,
`log d = 0` and the threshold is `0` in Lean. -/
theorem theorem_3_2 :
    ∃ c : ℝ, 0 < c ∧ ∀ (d n : ℕ), 1 ≤ n → (n : ℝ) * Real.log n ≤ c * Real.sqrt d →
      (∀ κ : Algorithm d n, IsMarkovKernel κ →
        PrivRobustStat.IsApproxDP (fun X => κ X) 1 (c / n) →
        1 / 4 ≤ experimentProb (Pdiag d) n κ
          (fun μ _ _ => {y | c * (d : ℝ) ^ ((3 : ℝ) / 2) / ((n : ℝ) ^ 2 * Real.log d) ≤ ‖y - μ‖})) ∧
      (∀ ε δ : ℝ, ε ≤ 1 → δ ≤ c / n →
        ENNReal.ofReal (c * (d : ℝ) ^ ((3 : ℝ) / 2) / ((n : ℝ) ^ 2 * Real.log d)) ≤
          bayesRisk (Pdiag d) n ε δ) := by sorry

end UnknownCovDP.General.Supplement
Source
Gaucher, Linder, Smith, Ullman, Zakynthinou, Dimension-Free Private Mean Estimation is Impossible Without the Covariance, STOC 2027 submission #573 (October 2026), https://github.com/sigact-stoc/stoc2027/releases/download/papers-1/573.pdf, Theorem 3.2, printed p. 16 (PDF p. 17); prior eq. (9) and Definition 3.1, same page; Bayes risk Definition 1.2, printed p. 2 (PDF p. 3)

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me