Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Bernoulli KL is antitone in its first argument below the target

Proved
BanditAlgorithm.bernoulliRelativeEntropy_antitone_fst

by Harry_Xu · Jul 29, 2026 · Mathlib 0df444a (Lean v4.33.1)

analysisbanditsrelative-entropy

Fix q∈(0,1)q\in(0,1)q∈(0,1). On the interval [0,q][0,q][0,q], Bernoulli relative entropy is nonincreasing in its first argument. Thus, whenever 0≤x≤y≤q0\le x\le y\le q0≤x≤y≤q,

d(y,q)≤d(x,q).d(y,q)\le d(x,q).d(y,q)≤d(x,q).

In words, moving the first Bernoulli parameter toward the fixed second parameter cannot increase their relative entropy.

This is the monotonicity step used in Lemma 10.8: if an empirical mean lies below μ+ε\mu+\varepsilonμ+ε, then its divergence from the larger threshold μ+Δ\mu+\Deltaμ+Δ is at least d(μ+ε,μ+Δ)d(\mu+\varepsilon,\mu+\Delta)d(μ+ε,μ+Δ).

Formalization Note: The first parameter xxx may equal zero. The proof rewrites the divergence using the continuous extension of xlog⁡xx\log xxlogx, proves the derivative is nonpositive in the interior, and then applies the derivative monotonicity theorem on the closed interval.

Preamble
import Definitions.Def_bernoulliRelativeEntropy

open Set
Formal statement
theorem BanditAlgorithm.bernoulliRelativeEntropy_antitone_fst
    {x y q : ℝ} (hx : x ∈ Set.Icc (0 : ℝ) 1)
    (hxy : x ≤ y) (hyq : y ≤ q) (hq : q ∈ Set.Ioo (0 : ℝ) 1) :
    BanditAlgorithm.bernoulliRelativeEntropy y q ≤
      BanditAlgorithm.bernoulliRelativeEntropy x q := by
  sorry
Source
Lattimore and Szepesvári, Bandit Algorithms (2020), monotonicity used in the proof of Lemma 10.8, printed pp. 138–139.

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