Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 3.1 — Witness encryption from bounded-factor MinRank

Proved
GenericGroupWE.Main.theorem_3_1

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

complexity-theorycryptographygeneric-groupsp2o-batch-stoc27ap2o-gran-per-chapterp2o-plan-paperp2o-strengthenedp2o-v1

Let n=max⁡(2,λ,∣φ∣)n=\max(2,\lambda,|\varphi|)n=max(2,λ,∣φ∣). Suppose a deterministic polynomial-time witness-preserving reduction computes 16n≤m=poly⁡(n)16n\le m=\operatorname{poly}(n)16n≤m=poly(n) and, for every supplied prime 2m≤p<2m+12^m\le p<2^{m+1}2m≤p<2m+1, an ordered linearly independent matrix basis. Valid witnesses efficiently produce a nonzero rank-one combination with Boolean right factor; on false statements every nonzero combination has rank at least the real number R(n)R(n)R(n). Assume R(n)→∞R(n)\to\inftyR(n)→∞ and R(n)log⁡2n=o(m)R(n)\log_2n=o(m)R(n)log2​n=o(m) uniformly over instances. There are polynomial-time oracle encryption and decryption algorithms, using the supplied prime, with correctness at least 1−2−n1-2^{-n}1−2−n on every instance and witness. For all sufficiently large nnn,

Q≤nR(n)/8⟹Adv⁡A≤n−R(n)/8Q\le n^{R(n)/8}\Longrightarrow\operatorname{Adv}_A\le n^{-R(n)/8}Q≤nR(n)/8⟹AdvA​≤n−R(n)/8

on every false statement, for every generic adversary of that query budget. Only security has a sufficiently-large threshold.

Differences from the source:

  • reading: m=poly(n) and R log n=o(m) although m also depends on lambda and phi → one uniform polynomial bound and an epsilon threshold valid for every instance of effective size at least that threshold; reason: this is the uniform reading needed by the pointwise guarantee and the application in §7
  • generalization: ordinary acyclic Boolean circuit instances → the published raw circuit evaluator, with missing or forward references reading zero; reason: ordinary circuits are included; total evaluation is still compiled gate by gate without any fan-out restriction
  • encoding: polynomial-time algorithms on encoded data → a single CookLevin polynomial-step implementation for each length-faithful encoded map; reason: the time predicate bounds actual Turing steps; clocks also bound rounds and state length
  • generalization: the theorem sentence gives p to encryption as a public input, and (1) obtains a group of that order → an added conclusion that every non-abort ciphertext carries a group context of order exactly p; reason: this is the first step of the construction and ties the existential algorithms to the supplied prime; dropping the clause gives the source claim
Preamble
import Definitions.Def_GenericGroupWE_Main_Reductions
Formal statement
namespace GenericGroupWE.Main

theorem theorem_3_1
    (r : RankReduction (ℕ × MIPRE.SAT.Circuit))
    (htime : r.PolyTimeOn circuitInstanceBits
      (fun i p => 2^(r.order i) ≤ p.val ∧ p.val < 2^(r.order i+1))
      (fun i w => SatisfiesCircuit i.2 w))
    (R : ℕ → ℝ)
    (hsize : ∀ i, 16 * circuitInstanceSize i ≤ r.order i)
    (hpoly : ∃ C d : ℕ, ∀ i, r.order i ≤ C * (circuitInstanceSize i + 1)^d)
    (hbasis : ∀ (i : ℕ × MIPRE.SAT.Circuit) (p : PrimeOrder),
      2^(r.order i) ≤ p.val → p.val < 2^(r.order i+1) →
      LinearIndependent (ZMod p.val) (r.basis i p).matrices ∧
      (∀ w, SatisfiesCircuit i.2 w → (r.witness i p w).Valid) ∧
      ((¬ ∃ w, SatisfiesCircuit i.2 w) → ∀ a,
        (r.basis i p).combination a ≠ 0 →
          R (circuitInstanceSize i) ≤ ((r.basis i p).combination a).rank))
    (hdiverges : ∀ B : ℝ, ∃ threshold : ℕ, ∀ n : ℕ, threshold ≤ n → B ≤ R n)
    (hlittleo : ∀ ε : ℝ, 0 < ε → ∃ threshold : ℕ,
      ∀ i : ℕ × MIPRE.SAT.Circuit, threshold ≤ circuitInstanceSize i →
        R (circuitInstanceSize i) * Real.logb 2 (circuitInstanceSize i : ℝ) ≤ ε * r.order i) :
    ∃ enc dec : OracleProgram,
      CookLevin.IsPolyTimeComputable enc.next ∧ CookLevin.IsPolyTimeComputable dec.next ∧
      (∀ (i : ℕ × MIPRE.SAT.Circuit) (p : PrimeOrder),
        2^(r.order i) ≤ p.val → p.val < 2^(r.order i+1) →
        (∀ message : Bool, ∀ ciphertext ∈ (suppliedEncryption enc i.1 i.2 p message).support,
          ciphertext.output.isSome → ∃ Γ, ciphertext.context = some Γ ∧ Γ.p = p.val) ∧
        ∀ w, SatisfiesCircuit i.2 w → ∀ message,
          1 - (2 : ℝ)^(-(circuitInstanceSize i : ℝ)) ≤ suppliedCorrectness enc dec i.1 i.2 p w message) ∧
      (∃ threshold : ℕ, ∀ (i : ℕ × MIPRE.SAT.Circuit), threshold ≤ circuitInstanceSize i →
        (¬ ∃ w, SatisfiesCircuit i.2 w) → ∀ p : PrimeOrder,
          2^(r.order i) ≤ p.val → p.val < 2^(r.order i+1) →
          ∀ Q : ℕ, (Q : ℝ) ≤ (circuitInstanceSize i : ℝ)^(R (circuitInstanceSize i)/8) →
            ∀ A : Adversary Q, suppliedAdvantage enc i.1 i.2 p A ≤
              (circuitInstanceSize i : ℝ)^(-(R (circuitInstanceSize i)/8))) := by sorry

end GenericGroupWE.Main
Source
Hair and Sahai, Witness Encryption via Prime-Order Generic Groups, STOC 2027 submission #47 (PDF built 2026-11-02; SHA-256 0b7f14e2db9bb3e199883ab907f854b7b5884c7c9f3d1e4c306bb71e89b659cb), https://github.com/sigact-stoc/stoc2027/releases/download/papers-1/47.pdf, Theorem 3.1, PDF p. 8 (printed p. 7).
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