Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Lemma 5.2 — HintSketch abstention and incorrect prediction

Proved
LinearSetReconciliation327.Main.lemma_5_2

by mikedeng1 · Nov 8, 2026 · Mathlib 0df444a (Lean v4.33.1)

linear-sketchp2o-batch-stoc27ap2o-gran-per-chapterp2o-plan-paperp2o-strengthenedp2o-v1set-reconciliationsparse-recovery

At a recursive or direct node, for a regular row (row hint JjJ_jJj​ and support SjS_jSj​ in [Nj][N_j][Nj​], ∣Jj∣≤BX|J_j|\le B_X∣Jj​∣≤BX​, yj=∣Sj∖Jj∣≤BYy_j=|S_j\setminus J_j|\le B_Yyj​=∣Sj​∖Jj​∣≤BY​), after a<ℓa<\ella<ℓ true recovered bits and over the row's coins: Pr⁡[worker returns ⊥]≤Ga(σj,yj+c)\Pr[\text{worker returns }\bot]\le G_a(\sigma_{j,y_j}+c)Pr[worker returns ⊥]≤Ga​(σj,yj​​+c) and Pr⁡[incorrect prediction]≤2−cGa(σj,yj+c)\Pr[\text{incorrect prediction}]\le2^{-c}G_a(\sigma_{j,y_j}+c)Pr[incorrect prediction]≤2−cGa​(σj,yj​​+c).

Differences from the source:

  • encoding: 0-based or 1-based coordinates [N] → 0-based Fin N or range N, with supports as finite sets and vector addition as symmetric difference; reason: the source identifies subsets with characteristic vectors and any injection relabels the coordinates
  • generalization: constants A, C fixed sufficiently large → all natural A, C; reason: the source proof uses only the row counts; the fixed constants are instances
Preamble
import Mathlib
import Definitions.Def_LinearSetReconciliation327_Main_Basic
import Definitions.Def_LinearSetReconciliation327_Main_Framework
import Definitions.Def_LinearSetReconciliation327_Main_HintSketch

open scoped ENNReal
open Finset
Formal statement
namespace LinearSetReconciliation327.Main

/-- Lemma 5.2, PDF p.21 (printed p.20): for a regular row of a non-leaf node, after `a < ℓ` recovered
bits of the true transcript. The probability is over the coins of the row. -/
theorem lemma_5_2 (P : HintParams) (N X Y : ℕ) (hargs : P.NodeArgs N X Y) (hnl : P.NonLeaf N X Y)
    (j : Fin (P.rows X Y)) (Jj Sj : Finset ℕ)
    (hJj : Jj ⊆ range (rowLen N (P.rows X Y) j.val)) (hSj : Sj ⊆ range (rowLen N (P.rows X Y) j.val))
    (hreg : ¬ P.Overloaded N X Y Jj Sj) (a : Fin (P.ell N X Y)) :
    let sch := P.scheme N X Y
    let σ : ℝ := (HintParams.sigma Jj.card (rowLen N (P.rows X Y) j.val) (Sj \ Jj).card : ℝ)
    let c : ℝ := (P.c N X Y : ℝ)
    let q : sch.Rand → Option (ZMod 2) := fun β =>
      sch.predict j β Jj (sch.truePrefix j β Sj a.val)
    prU (fun β : sch.Rand => q β = none) ≤ G a.val (σ + c) ∧
    prU (fun β : sch.Rand => ∃ b, q β = some b ∧ b ≠ sch.code j β Sj a) ≤
      ((2 : ℝ) ^ (P.c N X Y))⁻¹ * G a.val (σ + c) := by sorry

end LinearSetReconciliation327.Main
Source
Tianren Liu, Rui Xu, Tong Yang, Kangyang Zhou, STOC 2027 submission #327 (October 2026), Lemma 5.2, PDF p.21 (printed p.20)
Human review
  • Endorsed by Shuze Chen · Nov 8, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Nov 8, 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 paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me