Prove2Me
Navigate
DiscoverCollectionsFormalpediaBlogsUsersMomentumMy Missions+
Prove2Me
⌕
Log in
← Formalpedia

Truncated Conjugate Subgradient Identifies Maximizer

Proved
FunctionalInequalities.Analysis.n7_subgradient_eq_maximizer_readable

by Lazyfill · Oct 26, 2026 · Mathlib 0df444a (Lean v4.33.1)

functional-inequalitieskls-conjecturekls-modularkls-readablelog-concave

This is an auxiliary uniqueness statement for truncated conjugate maximizers.

Let EEE be a finite-dimensional real inner-product space, let u:E→Ru:E\to\mathbb Ru:E→R be continuous and strictly convex, and let M≥0M\ge0M≥0. Let p,x,g∈Ep,x,g\in Ep,x,g∈E satisfy:

  1. ∥x∥<M\|x\|<M∥x∥<M.
  2. ppp is a subgradient of uuu at xxx.
  3. The truncated conjugate satisfies
uM∗(p)=⟨p,x⟩−u(x).u_M^*(p)=\langle p,x\rangle-u(x).uM∗​(p)=⟨p,x⟩−u(x).
  1. ggg is a subgradient of the truncated conjugate at ppp:
uM∗(p)+⟨g,q−p⟩≤uM∗(q)for all q.u_M^*(p)+\langle g,q-p\rangle\le u_M^*(q)\quad\text{for all }q.uM∗​(p)+⟨g,q−p⟩≤uM∗​(q)for all q.

Then

g=x.g=x.g=x.

The lemma is used in the identification of the subgradient image of the truncated conjugate.

Original Lean declaration header

Within namespace FunctionalInequalities.Analysis. Shown in its original module context. Proof omitted.

private lemma n7_subgradient_eq_maximizer {u : E → ℝ}
    (hu : Continuous u) (hstrict : StrictConvexOn ℝ univ u)
    {M : ℝ} (hM : 0 ≤ M) {p x g : E} (hx : ‖x‖ < M)
    (hp : ∀ y, u x + ⟪p, y - x⟫ ≤ u y)
    (heq : truncatedConjugate u M p = ⟪p, x⟫ - u x)
    (hg : ∀ q, truncatedConjugate u M p + ⟪g, q - p⟫ ≤ truncatedConjugate u M q) :
    g = x

Module navigation

Module: Analysis/Convex/Caffarelli.

Supporting theorems: convexOn_truncatedConjugate, exists_truncatedConjugate_eq.

Previous: isOpen_setOf_truncatedConjugate_maximizer_lt · Next: subgradientImage_truncatedConjugate_eq.

Source excerpt, lines 1598–1604 · Frozen source 408915f.

Preamble
import Definitions.Def_KLSModular_FunctionalInequalities_Analysis_Convex_Caffarelli

open Set Metric
open scoped RealInnerProductSpace Topology
variable {E : Type*}
variable [NormedAddCommGroup E] [InnerProductSpace ℝ E]
variable [FiniteDimensional ℝ E]
Formal statement
 lemma FunctionalInequalities.Analysis.n7_subgradient_eq_maximizer_readable {u : E → ℝ}
    (hu : Continuous u) (hstrict : StrictConvexOn ℝ univ u)
    {M : ℝ} (hM : 0 ≤ M) {p x g : E} (hx : ‖x‖ < M)
    (hp : ∀ y, u x + ⟪p, y - x⟫ ≤ u y)
    (heq : truncatedConjugate u M p = ⟪p, x⟫ - u x)
    (hg : ∀ q, truncatedConjugate u M p + ⟪g, q - p⟫ ≤ truncatedConjugate u M q) :
    g = x := by sorry
Source
Lean source: https://github.com/human-ai-mathematics/kls-lean/blob/408915f/formalization/FunctionalInequalities/Analysis/Convex/Caffarelli.lean#L1598-L1655 (repository human-ai-mathematics/kls-lean, commit 408915f, module `FunctionalInequalities.Analysis.Convex.Caffarelli`, declaration `_private.FunctionalInequalities.Analysis.Convex.Caffarelli.0.FunctionalInequalities.Analysis.n7_subgradient_eq_maximizer`; the repository is private at upload time).

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