Local Chain Rule for the Diffusion Generator
ProvedFunctionalInequalities.Probability.Diffusion.diffusionGenerator_comp_of_contDiffAt_modularfunctional-inequalitieskls-conjecturekls-modularlog-concave
This is the local, pointwise form of the chain rule for diffusionGenerator.
Let and be real normed additive groups, let , let , , , , and . Assume only that is at and is at , namely ContDiffAt ℝ 2 Φ (G x) and ContDiffAt ℝ 2 G x.
With ,
This local variant avoids global smoothness assumptions and can be reused whenever only the behavior at one state point is available.
Superseded upload. This snapshot was replaced by FunctionalInequalities.Probability.Diffusion.diffusionGenerator_comp_of_contDiffAt_readable: open the readable replacement. The replacement preserves the frozen source and mathematical statement, with a compact declaration context. This historical node remains available for existing imports.
Preamble
import Definitions.Def_KLSModular_FunctionalInequalities_Probability_Diffusion_Generator
noncomputable section
section
namespace FunctionalInequalities
end FunctionalInequalities
open FunctionalInequalities
section
namespace FunctionalInequalities.Probability
end FunctionalInequalities.Probability
open FunctionalInequalities FunctionalInequalities.Probability
section
namespace FunctionalInequalities.Probability.Diffusion
end FunctionalInequalities.Probability.Diffusion
open FunctionalInequalities FunctionalInequalities.Probability FunctionalInequalities.Probability.Diffusion
section S3B1LLocalChainRule
section NormedSpace
variable {V W : Type*}
variable [NormedAddCommGroup V] [NormedSpace ℝ V] [NormedAddCommGroup W] [NormedSpace ℝ W]
Formal statement
theorem FunctionalInequalities.Probability.Diffusion.diffusionGenerator_comp_of_contDiffAt_modular {d : ℕ} (b : V → V) (σ : V → Fin d → V)
(Φ : W → ℝ) (G : V → W) (x : V) (hΦ : ContDiffAt ℝ 2 Φ (G x)) (hG : ContDiffAt ℝ 2 G x) :
diffusionGenerator b σ (Φ ∘ G) x =
fderiv ℝ Φ (G x) (fderiv ℝ G x (b x) +
(1 / 2 : ℝ) • ∑ k, fderiv ℝ (fderiv ℝ G) x (σ x k) (σ x k)) +
(1 / 2 : ℝ) * ∑ k, fderiv ℝ (fderiv ℝ Φ) (G x) (fderiv ℝ G x (σ x k)) (fderiv ℝ G x (σ x k)) := by sorry
Source
Lean source: https://github.com/human-ai-mathematics/kls-lean/blob/408915f/formalization/FunctionalInequalities/Probability/Diffusion/Generator.lean#L240-L298 (repository human-ai-mathematics/kls-lean, commit 408915f, module `FunctionalInequalities.Probability.Diffusion.Generator`, declaration `FunctionalInequalities.Probability.Diffusion.diffusionGenerator_comp_of_contDiffAt`; the repository is private at upload time).