Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

A short mixed bank has a translation or reflected collision

Proved
Couret.STROBO2TORS.recoh_short_bank_collision_dichotomy

by alexcour · Nov 8, 2026 · Mathlib c5ea003 (Lean v4.30.0)

couret-programmecouret-unification-20261108

Let GGG be an abelian group and RRR a commutative ring. Let finite abelian groups H,VH,VH,V embed additively and injectively into GGG, with 2H=02H=02H=0, 3V=03V=03V=0 and ∣V∣=3s|V|=3^s∣V∣=3s. Let r>0r>0r>0, k<r+sk<r+sk<r+s, and functions fi:G→Rf_i:G\to Rfi​:G→R for i<ki<ki<k satisfy fi(x+y)=fi(x)fi(y)f_i(x+y)=f_i(x)f_i(y)fi​(x+y)=fi​(x)fi​(y) and fi(0)=1f_i(0)=1fi​(0)=1. Supply additive bi:H→Z/2Zb_i:H\to\mathbb Z/2\mathbb Zbi​:H→Z/2Z and ci:V→Z/3Zc_i:V\to\mathbb Z/3\mathbb Zci​:V→Z/3Z such that on embedded HHH, fi=1f_i=1fi​=1 when bi=0b_i=0bi​=0 and fi=−1f_i=-1fi​=−1 otherwise; on embedded VVV, ci=0c_i=0ci​=0 implies fi=1f_i=1fi​=1. Assume that injectivity of the full binary map implies existence of rrr distinct selected indices whose binary map onto (Z/2Z)r(\mathbb Z/2\mathbb Z)^r(Z/2Z)r is surjective. For every triple S={x,y,z}S=\{x,y,z\}S={x,y,z}, one of the following collisions exists:

  1. A nonzero ddd with 2d=02d=02d=0 such that S≠S+dS\ne S+dS=S+d, their ordered squares agree, and all kkk probe sums agree.
  2. Nonzero d,vd,vd,v with 2d=02d=02d=0, 3v=03v=03v=0 such that A={0,v,v+d}A=\{0,v,v+d\}A={0,v,v+d} and B={0,−v,−v+d}B=\{0,-v,-v+d\}B={0,−v,−v+d} are distinct, their ordered squares agree, and all kkk probe sums agree.
Translation collision∨Reflected collision.\text{Translation collision}\quad\lor\quad\text{Reflected collision}.Translation collision∨Reflected collision.

No assumption ∣H∣=2r|H|=2^r∣H∣=2r is made directly; the stated information-set contract supplies the needed rank property.

Preamble
import Aesop
import Definitions.Def_Couret_RecohRankKernel_9b83850e485b
import Lean.Elab.Tactic.Omega
import Mathlib.Algebra.BigOperators.Group.Finset.Basic
import Mathlib.Data.Finset.Card
import Mathlib.Data.Finset.Powerset
import Mathlib.Data.Fintype.BigOperators
import Mathlib.Data.Fintype.Card
import Mathlib.Data.ZMod.Basic
import Mathlib.Tactic.Abel
import Mathlib.Tactic.Ring
import Mathlib.Tactic.Tauto
import Theorems.Thm_Couret_STROBO2TORS_recoh_binary_bank_kernel_or_injective
import Theorems.Thm_Couret_STROBO2TORS_recoh_binary_kernel_genuine_collision
import Theorems.Thm_Couret_STROBO2TORS_recoh_pivot_bank_has_genuine_collision
import Definitions.Def_Couret_RecohRankKernel_9b83850e485b

set_option autoImplicit false

/-!
RECOH: binary evaluation dichotomy and exact collision reduction.

This module eliminates a manual case split for arbitrary finite banks:
either a nonzero binary kernel exists or the binary evaluation is injective.
In the injective case it records the information-set selection as one
EXPLICIT hypothesis. It does not claim that selection for general rank.

The rank-one case has a separate certified branch and is not duplicated.
-/

namespace Couret.STROBO2TORS

variable {G : Type*} [AddCommGroup G] [DecidableEq G]





Formal statement

theorem recoh_short_bank_collision_dichotomy
    {H V R : Type*}
    [AddCommGroup H] [Fintype H]
    [AddCommGroup V] [Fintype V]
    [CommRing R]
    (r s k : ℕ) (hr : 0 < r) (hk : k < r + s)
    (hcardV : Fintype.card V = 3 ^ s)
    (h2 : ∀ a : H, a + a = 0)
    (h3 : ∀ a : V, (a + a) + a = 0)
    (embH : H →+ G) (injH : Function.Injective embH)
    (embV : V →+ G) (injV : Function.Injective embV)
    (bank : Fin k → G → R)
    (hadd : ∀ i : Fin k, ∀ a b : G,
      bank i (a + b) = bank i a * bank i b)
    (hval0 : ∀ i : Fin k, bank i 0 = 1)
    (binary : Fin k → H →+ ZMod 2)
    (hbinary0 : ∀ i : Fin k, ∀ a : H,
      binary i a = 0 → bank i (embH a) = 1)
    (hbinary1 : ∀ i : Fin k, ∀ a : H,
      binary i a ≠ 0 → bank i (embH a) = -1)
    (ternary : Fin k → V →+ ZMod 3)
    (hternary : ∀ i : Fin k, ∀ a : V,
      ternary i a = 0 → bank i (embV a) = 1)
    (hinfo : Function.Injective
        (fun a : H => fun i : Fin k => binary i a) →
      ∃ pivot : Fin r → Fin k,
        Function.Injective pivot ∧
        Function.Surjective
          (fun a : H => fun j : Fin r => binary (pivot j) a))
    (x y z : G) (hcard : ({x, y, z} : Finset G).card = 3) :
    (∃ d : G, d ≠ 0 ∧ d + d = 0 ∧
      ({x, y, z} : Finset G) ≠ {x + d, y + d, z + d} ∧
      (∀ q : G,
        orderedSquare ({x, y, z} : Finset G) q =
          orderedSquare ({x + d, y + d, z + d} : Finset G) q) ∧
      (∀ i : Fin k,
        recohTripleProbe (bank i) x y z =
          recohTripleProbe (bank i) (x + d) (y + d) (z + d))) ∨
    (∃ d v : G, d ≠ 0 ∧ d + d = 0 ∧
      v ≠ 0 ∧ (v + v) + v = 0 ∧
      ({0, v, v + d} : Finset G) ≠ {0, -v, -v + d} ∧
      (∀ q : G,
        orderedSquare ({0, v, v + d} : Finset G) q =
          orderedSquare ({0, -v, -v + d} : Finset G) q) ∧
      (∀ i : Fin k,
        recohTripleProbe (bank i) 0 v (v + d) =
          recohTripleProbe (bank i) 0 (-v) (-v + d))) := by sorry


end Couret.STROBO2TORS

Source
Alexandre Couret, Couret — Additive reconstruction: STROBO-4, singleton collisions and RECOH formal proof artifacts, https://zenodo.org/records/23980513; archive COURET_MATHEMATIQUES_20261105.zip, SHA-256 19a3b59aee930df74cab79c27b29e014843bd17179a26fef1261ed6b5a651ab7; file RECOH/lean/RecohBinaryDichotomy.lean, lines 97–157; source file SHA-256 740733aa832d9a2d29988c8cbc79b0861bfba8e7e0e0aa8e9f9826db98ef1779; theorem Couret.STROBO2TORS.recoh_short_bank_collision_dichotomy

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