Provedbalanced-conversionconversion-theoremattention-compute-lawlower-bound
Uniform proposal inside a block of m keys with proposal mass ρ contributes Σπ²/(ρ/m) ≤ m (Σπ)²/ρ
For all , , and , if , and (given ), then
What the notation means · or point at any symbol
- Type
- Type: a definition from the library, not special notation. Its name is its meaning; open it in Lean to see how it is defined. (
Type) - Fintype
- Fintype: a definition from the library, not special notation. Its name is its meaning; open it in Lean to see how it is defined. (
Fintype) - f(x)Function application
- Fintype applied to J. Lean writes f x; papers write f(x). (
Fintype J) - ⟹Implies
- If J, then ℝ. (Between types, A → B is the type of functions from A to B.) (
J → ℝ) - NNatural numbers
- The natural numbers 0, 1, 2, …. In Lean they start at 0. (
ℕ) - ∀For all
- The statement after the comma holds for every j. (
∀ j, 0 ≤ π j) - <Less than
- 0 is strictly smaller than ρ. (
0 < ρ) - ∑Summation
- Add up π(j)^2/(ρ/m) for every j. (
∑ j, π j ^ 2 / (ρ / m)) - RReal numbers
- The real numbers: every point on the number line. (
ℝ) - ≤At most
- 0 is less than or equal to π(j). (
0 ≤ π j) - aᵇPower
- π(j) raised to the power 2. (
π j ^ 2) - πPi
- The ratio of a circle's circumference to its diameter, ≈ 3.14159. (
π) - a/bDivision
- ρ divided by m. In Lean, dividing by 0 gives 0. (
ρ / m) - ·Multiplication
- m times (∑ j, π(j))^2. (
m * (∑ j, π j) ^ 2)
theorem bc2-uniform-block-contribution {J : Type} [Fintype J] (π : J → ℝ) (ρ : ℝ) (m : ℕ) (hπ : ∀ j, 0 ≤ π j) (hρ : 0 < ρ) (hm : 0 < m) :
∑ j, π j ^ 2 / (ρ / m) ≤ m * (∑ j, π j) ^ 2 / ρWith block mass ≤ n^{−3} and ρ ≥ (block mass)/K the truncated blocks contribute at most K m n^{−3} each, O(1) in total; closes the truncation step of the n^{O(ε²)} bound. Part of Balanced Conversion: the Conversion Theorem's two-way descent factor is exactly 1 + (p−q)²/(q(1−q)), so additive oracle error is second order at mass-balanced nodes; a mass-balanced descent converts any deterministic additive-ε oracle with loss n^{O(ε²)}, and the biased product family shows this is tight. Quantifies the last residual class of the Attention Compute Law.
How it was proved
- Proved
- 4 October 2026
- By
- Ajax, with the agent “Baguette math & proofs (Claude Code, 2026-10-03)”
- Checked by
- Lean's kernel (Lean 4.33.1, with Mathlib); the proof is signed and replays offline
- Axioms
- Classical.choice, Quot.sound, propext
- Proof
- 19 lines of Lean
Sign in to see the proofThe proof, its signed report and its history.