Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Sharp exact-rank Gaussian posterior-overshoot sum

Proved
BanditAlgorithm.gaussian_ts_suboptimal_exact_rank_overshoot_sum_sharp

by Harry_Xu · Jul 31, 2026 · Mathlib 0df444a (Lean v4.33.1)

asymptoticsbanditsgaussianthompson-sampling

Fix an optimal arm i0i_0i0​, a suboptimal arm iii of gap Δi>0\Delta_i>0Δi​>0, and 0<ε<Δi0<\varepsilon<\Delta_i0<ε<Δi​. Along the exact realized ranks of arm iii, the expected number of posterior distributions that put more than 1/n1/n1/n mass above μ∗−ε\mu_* - \varepsilonμ∗​−ε satisfies

E ⁣[∑s<Ti(n)1 ⁣{Gi,s(μ∗−ε)>1n}]≤1+2(Δi−ε)2(log⁡n+πlog⁡n+1).\mathbb E\!\left[\sum_{s<T_i(n)}\mathbf 1\!\left\{G_{i,s}(\mu_* - \varepsilon)>\frac1n\right\}\right] \le 1+\frac{2}{(\Delta_i-\varepsilon)^2}\left(\log n+\sqrt{\pi\log n}+1\right).E​s<Ti​(n)∑​1{Gi,s​(μ∗​−ε)>n1​}​≤1+(Δi​−ε)22​(logn+πlogn​+1).

The leading coefficient is therefore the sharp 2/(Δi−ε)22/(\Delta_i-\varepsilon)^22/(Δi​−ε)2 from Exercise 36.6(b).

Preamble
import Definitions.Def_ThompsonSampling
import Definitions.Def_GaussianBandit

open MeasureTheory ProbabilityTheory ENNReal
Formal statement
theorem BanditAlgorithm.gaussian_ts_suboptimal_exact_rank_overshoot_sum_sharp :
      ∀ (k : ℕ) [NeZero k] (μvec : Fin k → ℝ) (π : BanditAlgorithm.BanditPolicy k),
        ∀ (i₀ : Fin k),
          BanditAlgorithm.banditArmMean (BanditAlgorithm.gaussianBandit μvec) i₀ =
            BanditAlgorithm.banditOptimalMean (BanditAlgorithm.gaussianBandit μvec) →
        ∀ (i : Fin k), 0 < BanditAlgorithm.banditGap (BanditAlgorithm.gaussianBandit μvec) i →
        ∀ ε : ℝ, 0 < ε →
          ε < BanditAlgorithm.banditGap (BanditAlgorithm.gaussianBandit μvec) i →
        ∀ n : ℕ, 2 ≤ n →
          (∫⁻ h, ∑ s ∈ Finset.range (BanditAlgorithm.armPullCount i h),
              (if 1 / (n : ℝ) <
                  BanditAlgorithm.gaussianTSTailProb i s
                    (BanditAlgorithm.banditArmMean (BanditAlgorithm.gaussianBandit μvec) i₀ - ε) h
                then (1 : ℝ≥0∞) else 0)
              ∂BanditAlgorithm.banditMeasure (BanditAlgorithm.gaussianBandit μvec) π n) ≤
            ENNReal.ofReal
              (1 + 2 / (BanditAlgorithm.banditGap (BanditAlgorithm.gaussianBandit μvec) i - ε) ^ 2 *
                (Real.log n + Real.sqrt (Real.pi * Real.log n) + 1)) := by sorry
Source
Lattimore and Szepesvári, Bandit Algorithms (CUP 2020), Exercise 36.6(b), printed p. 475 / PDF p. 484, and Theorem 36.2, Eq. (36.3), printed pp. 463–465 / PDF pp. 472–474.

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