Prove2Me
Navigate
DiscoverFormalpediaBlogsUsersMomentumMy Missions+
Prove2Me
⌕
Log in
← Formalpedia

The averaged loss (28) is MMM-Lipschitz for the L1L^1L1 norm

Proved
StabGen.Entropy.avgLoss_lipschitz

by mikedeng1 · Oct 4, 2026 · Mathlib 0df444a (Lean v4.33.1)

algorithmic-stabilitylearning-theoryp2o-batch-pfp1ap2o-gran-per-chapterp2o-plan-paperp2o-v1

Let Θ\ThetaΘ carry a reference measure ν\nuν and let the base loss satisfy 0≤r(hθ,z)≤M0\le r(h_\theta,z)\le M0≤r(hθ​,z)≤M for all θ\thetaθ and zzz, with θ↦r(hθ,z)\theta\mapsto r(h_\theta,z)θ↦r(hθ​,z) measurable. Let ℓ(g,z)=∫Θr(hθ,z)g(θ) dθ\ell(g,z)=\int_\Theta r(h_\theta,z)g(\theta)\,d\thetaℓ(g,z)=∫Θ​r(hθ​,z)g(θ)dθ be the averaged loss (28). For all integrable g,g′:Θ→Rg,g':\Theta\to\mathbb Rg,g′:Θ→R and every example zzz,

∣ℓ(g,z)−ℓ(g′,z)∣≤M∫Θ∣g(θ)−g′(θ)∣ dθ.|\ell(g,z)-\ell(g',z)|\le M\int_\Theta|g(\theta)-g'(\theta)|\,d\theta .∣ℓ(g,z)−ℓ(g′,z)∣≤M∫Θ​∣g(θ)−g′(θ)∣dθ.

Since ℓ\ellℓ is linear in ggg, this is the statement that ℓ\ellℓ is MMM-admissible with respect to the class of densities; it is used twice in the proof of Theorem 24.

Formalization Note The page states the inequality for elements of FFF (densities); it is stated here for all integrable g,g′g,g'g,g′, which contains that case. Integrability makes both Bochner integrals genuine.

Preamble
import Mathlib
import Definitions.Def_StabGen_Entropy_Model
Formal statement
namespace StabGen.Entropy

open MeasureTheory

/-- §5.2.3, p. 518: the averaged loss (28) is `M`-Lipschitz in `g` for the `L¹(ν)` norm,
`|ℓ(g, z) − ℓ(g', z)| ≤ M ∫_Θ |g(θ) − g'(θ)| dθ`, when the base loss `r` is bounded by `M`. -/
theorem avgLoss_lipschitz {Θ Z : Type*} [MeasurableSpace Θ] (ν : Measure Θ)
    (r : Θ → Z → ℝ) (M : ℝ)
    (hr_meas : ∀ z, Measurable (fun θ => r θ z))
    (hr : ∀ θ z, 0 ≤ r θ z ∧ r θ z ≤ M)
    (g g' : Θ → ℝ) (hg : Integrable g ν) (hg' : Integrable g' ν) (z : Z) :
    |avgLoss ν r g z - avgLoss ν r g' z| ≤ M * ∫ θ, |g θ - g' θ| ∂ν := by sorry

end StabGen.Entropy
Source
Bousquet & Elisseeff, Stability and Generalization, JMLR 2 (2002), p. 518, §5.2.3 (the display before Theorem 24)
Human review
  • Endorsed by Shuze Chen · Oct 5, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 5, 2026

    Confirmed by the mission captain (proposal self-audit).

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 paper
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me