PFR/EntropyPFR.lean
Explanation legend Lean4 Domain Explanation Language (L4DEL)
L4DEL explanations are a separate layer. They do not change the Lean source. Intuition is explanatory prose; numbered items are a reading path; sources link the wording back to evidence.
- L4DEL
- Lean4 Domain Explanation Language
Credit: The upstream authors and contributors retain credit for the project and mathematics. lean4-explainer only renders their pinned source and adds the separately disclosed explanation layer.
Project sources: Pinned project documentation, Polynomial Freiman-Ruzsa proof paper, arXiv v2
2 declarations, 0 contain sorry, 0 declared as axioms
Entropic version of polynomial Freiman-Ruzsa conjecture
Here we prove the entropic version of the polynomial Freiman-Ruzsa conjecture.
Main results
entropic_PFR_conjecture: For two $G$-valued random variables $X_1^0, X_2^0$, there is some subgroup $H \leq G$ such that $d[X_1^0;U_H] + d[X_2^0;U_H] \le 11 d[X_1^0;X_2^0]$, and both $d[X_1^0;U_H]$ and $d[X_2^0;U_H]$ are at most $6 d[X_1^0;X_2^0]$.
Imported by 2 files
- theorem
tau_strictly_decreasesIf $d[X_1;X_2] > 0$ then there are $G$-valued random variables $X'_1, X'_2$ such that $\tau[X'_1;X'_2] < \tau[X_1;X_2]$ line 39 sourceSource documentation
If $d[X_1;X_2] > 0$ then there are $G$-valued random variables $X'_1, X'_2$ such that $\tau[X'_1;X'_2] < \tau[X_1;X_2]$. Phrased in the contrapositive form for convenience of proof.
standard axioms: Classical.choice, Quot.sound, propext what this means
Statement uses:
MeasureTheory.MeasureSpace,MeasureTheory.IsProbabilityMeasure,MeasureTheory.MeasureSpace.volume,AddCommGroup,Module,ZMod,OfNat.ofNat,Nat,Finite,MeasurableSpace,MeasurableSingletonClass, refPackage,Measurable, TauMinimizes,Eq,Real,refPackage.η,HDiv.hDiv, rdistProof uses:
Fintype,Nonempty,nonempty_fintype,Nonempty.intro,Exists,MeasureTheory.Measure,And,ProbabilityTheory.iIndepFun,Fin,Nat.succ,Matrix.vecCons,Matrix.vecEmpty,ProbabilityTheory.IdentDistrib, ProbabilityTheory.independent_copies4_nondep,Eq.mpr,id,congrArg,Eq.symm, ProbabilityTheory.IdentDistrib.rdist_congr,MeasureTheory.MeasureSpace.mkand 7 more36 instances, coercions and auxiliary constants left out.
Tactic trace: 7 steps
Open to load the checked before-and-after goal states.
theorem tau_strictly_decreases (h_min : TauMinimizes p X₁ X₂) (hpη : p.η = 1 / 9) : d[X₁ # X₂] = 0 := by …Mentions: TauMinimizes
- theorem
entropic_PFR_conjectureEntropic PFR (constants 11 and 6) line 54 sourceSource documentation
entropic_PFR_conjecture: For two $G$-valued random variables $X_1^0, X_2^0$, there is some subgroup $H \leq G$ such that $d[X_1^0;U_H] + d[X_2^0;U_H] \le 11 d[X_1^0;X_2^0]$, and both $d[X_1^0; U_H]$ and $d[X_2^0; U_H]$ are at most $6 d[X_1^0;X_2^0]$.L4DEL explanation
Any two random variables in an F2-vector space are within a constant multiple of their Ruzsa distance of a uniform distribution on some subgroup.
- Statement: for finite G of characteristic 2 and a package of two G-valued variables X01, X02 with parameter eta = 1/9, there is a subspace H and a uniform variable U on H with d[X01 ; U] + d[X02 ; U] at most 11 d[X01 ; X02], and each of d[X01 ; U] and d[X02 ; U] at most 6 d[X01 ; X02].
- Take a pair X1, X2 minimizing the tau functional (tau_minimizer_exists). The theorem tau_strictly_decreases then forces d[X1 ; X2] = 0 when eta = 1/9.
- Zero Ruzsa distance means X1 and X2 are uniform on cosets of a common subgroup (exists_isUniform_of_rdist_eq_zero), which gives H and U.
- The distances from the original variables to U follow from minimality of tau and the Ruzsa triangle inequality (rdist_triangle), finished by linear arithmetic.
Prerequisites: tau_minimizer_exists, tau_strictly_decreases, rdist_triangle, exists_isUniform_of_rdist_eq_zero
See also: entropic_PFR_conjecture_improv
Sources and attribution: Pinned Lean source (ce7ac95e620a)
standard axioms: Classical.choice, Quot.sound, propext what this means
Statement uses:
MeasureTheory.MeasureSpace,MeasureTheory.IsProbabilityMeasure,MeasureTheory.MeasureSpace.volume,AddCommGroup,Module,ZMod,OfNat.ofNat,Nat,Finite,MeasurableSpace,MeasurableSingletonClass, refPackage,Eq,Real,refPackage.η,HDiv.hDiv,Exists,Submodule,And,Measurableand 7 moreProof uses:
Fintype,Nonempty,nonempty_fintype,Nonempty.intro, TauMinimizes, tau_minimizer_exists, tau_strictly_decreases,AddSubgroup, exists_isUniform_of_rdist_eq_zero,Prod, rdist_symm, tau, is_tau_min,refPackage.hmeas2,refPackage.hmeas1, rdist_triangle,Exists.intro,OrderIso,AddSubgroup.toZModSubmodule,inferInstanceand 97 more135 instances, coercions and auxiliary constants left out.
tau functional: tau[X1 ; X2] = d[X1 ; X2] + eta d[X01 ; X1] + eta d[X02 ; X2], where X01 and X02 are the fixed reference variables of the package; minimizing it is the engine of the entropic proof.
Tactic trace: 16 steps
Open to load the checked before-and-after goal states.
theorem entropic_PFR_conjecture (hpη : p.η = 1 / 9) : ∃ H : Submodule (ZMod 2) G, ∃ Ω : Type uG, ∃ mΩ : MeasureSpace Ω, ∃ U : Ω → G, IsProbabilityMeasure (ℙ : Measure Ω) ∧ Measurable U ∧ IsUniform H U ∧ d[p.X₀₁ # U] + d[p.X₀₂ # U] ≤ 11 * d[p.X₀₁ # p.X₀₂] ∧ d[p.X₀₁ # U] ≤ 6 * d[p.X₀₁ # p.X₀₂] ∧ d[p.X₀₂ # U] ≤ 6 * d[p.X₀₁ # p.X₀₂] := by …