Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Lemma 8 bridge — Gittins tie rules realize optimal cost extremes

Proved
PandoraBO.Private20261023.optimal_face_gittins_cost_extrema

by QianJaneXie · Oct 24, 2026 · Mathlib 0df444a (Lean v4.33.1)

bayesian-optimizationexpected-budgetgittins-indexpandoras-box

Fix λ>0\lambda>0λ>0. There exist two Gittins policies π−\pi_-π−​ and π+\pi_+π+​ whose expected costs bound the costs of every penalized-optimal policy:

C(π−)≤C(ρ)≤C(π+)(ρ optimal at λ).C(\pi_-)\le C(\rho)\le C(\pi_+)\qquad(\rho\text{ optimal at }\lambda).C(π−​)≤C(ρ)≤C(π+​)(ρ optimal at λ).

For every a∈[0,1]a\in[0,1]a∈[0,1], there also exists a Gittins policy with expected cost

aC(π−)+(1−a)C(π+).aC(\pi_-)+(1-a)C(\pi_+).aC(π−​)+(1−a)C(π+​).

By Theorem 1, the two witnesses and the interpolating policies are themselves penalized-optimal. This is an additional bridge required by the budget argument; it is not claimed to be a verbatim numbered lemma in the paper. It makes explicit both extreme-cost attainment within the Gittins class and randomization through admissible ties.

Preamble
import Definitions.Def_PandoraBO_Model

set_option autoImplicit false
Formal statement
namespace PandoraBO.Private20261023

theorem optimal_face_gittins_cost_extrema {n : ℕ} (M : Instance n)
    (multiplier : ℝ) (hpos : 0 < multiplier) :
    ∃ low high : Policy n,
      isGittinsPolicy M multiplier low ∧ isGittinsPolicy M multiplier high ∧
      (∀ π : Policy n, optimal M multiplier π →
        expectedCost M low ≤ expectedCost M π ∧ expectedCost M π ≤ expectedCost M high) ∧
      ∀ a : ℝ, 0 ≤ a → a ≤ 1 → ∃ π : Policy n,
        isGittinsPolicy M multiplier π ∧
        expectedCost M π = a * expectedCost M low + (1 - a) * expectedCost M high := by sorry

end PandoraBO.Private20261023
Source
Qian Xie, Raul Astudillo, Peter I. Frazier, Ziv Scully, and Alexander Terenin. [Cost-aware Bayesian Optimization via the Pandora's Box Gittins Index](https://arxiv.org/abs/2406.20062). NeurIPS 2024. Appendix B.5, Lemma 8 and the conclusion of Theorem 2. This is a supporting formalization lemma for expected-cost extrema among optimal Gittins policies, not a separately numbered theorem in the paper.
Human review
  • Endorsed by Shuze Chen · Oct 24, 2026

    Confirmed by the moderator at approval.

  • Endorsed by QianJaneXie · Oct 24, 2026

    Confirmed by the mission captain (proposal self-audit).

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