Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Lemma G.7 — the comparison law built from a centered allocation is feasible in program (3.6)

Proved
IAFTRL.lemma_G_7

by Canzhe Zhao · Nov 17, 2026 · Mathlib 0df444a (Lean v4.33.1)

bandit-algorithmslinear-banditsonline-learning

Setting. Fix a round n≥1n\ge1n≥1 of GIA-FTRL (Algorithm 2) with cumulative estimate SSS. The objects are:

  • K=Conv B\mathcal K=\mathrm{Conv}\,\mathcal BK=ConvB, a compact generator B\mathcal BB, and Δ(a)=(a−a∗)⊤θ\Delta(a)=(a-a^*)^\top\thetaΔ(a)=(a−a∗)⊤θ with 0≤Δ≤20\le\Delta\le20≤Δ≤2 on K\mathcal KK and a∗∈Ka^*\in\mathcal Ka∗∈K;
  • the exponential-weights law qn∝e−x⊤S/βnμ(dx)q_n\propto e^{-x^\top S/\beta_n}\mu(dx)qn​∝e−x⊤S/βn​μ(dx) and π∗=μ{a∗}>0\pi^*=\mu\{a^*\}>0π∗=μ{a∗}>0;
  • a finite feasible centered sampling allocation ndesn_{\rm des}ndes​ on B\mathcal BB (1.4), with total mass NNN.

For L≥1L\ge1L≥1 put ρ=16BL2/n\rho=16BL^2/nρ=16BL2/n and define the comparison law (Definition G.2)

pc=(1−γ) qn+ρ ndes∣{a≠a∗},γ=ρ ndes({a≠a∗}).p^c=(1-\gamma)\,q_n+\rho\,n_{\rm des}|_{\{a\ne a^*\}},\qquad\gamma=\rho\,n_{\rm des}(\{a\ne a^*\}).pc=(1−γ)qn​+ρndes​∣{a=a∗}​,γ=ρndes​({a=a∗}).

Claim. Assume the following, which are the properties of the comparison event WtW_tWt​ used in the source:

  1. the reference action of (3.4) is a∗a^*a∗;
  2. X(a)=(a−a∗)⊤S≥0X(a)=(a-a^*)^\top S\ge0X(a)=(a−a∗)⊤S≥0 on K\mathcal KK;
  3. Δ(a)≤L X(a)/n\Delta(a)\le L\,X(a)/nΔ(a)≤LX(a)/n on K\mathcal KK;
  4. ρ≤1\rho\le1ρ≤1 and ρN≤min⁡{Γn,1/4}\rho N\le\min\{\Gamma_n,1/4\}ρN≤min{Γn​,1/4};
  5. qn(a∗)≥3/4q_n(a^*)\ge3/4qn​(a∗)≥3/4;
  6. 64∫a≠a∗e−X/(2βn)dμ≤2dρ2π∗64\int_{a\ne a^*}e^{-X/(2\beta_n)}d\mu\le2d\rho^2\pi^*64∫a=a∗​e−X/(2βn​)dμ≤2dρ2π∗.

Then pcp^cpc is feasible in program (3.6) at round nnn: the mixture budget, positive-definite design, update-magnitude, second-moment and proxy constraints all hold.

Formalization Note. This is Lemma G.7 in explicit-hypothesis form: the event WtW_tWt​ is replaced by the consequences of its definition and of Lemma G.3 that the proof uses.

Preamble
import Mathlib
import Definitions.Def_IAFTRL_LinEnv
import Definitions.Def_IAFTRL_GIAFTRL
Formal statement
namespace IAFTRL

open MeasureTheory

/-- **Lemma G.7 (The comparison distribution constructed from a centered sampling allocation
satisfies all budgets).** Zhao–Tan, *Best-of-Both-Worlds FTRL for Linear Bandits* (STOC submission version), Lemma G.7, pp.134–137: "For `t ≥ T*`, on `W_t`, the
distribution `p^c_t = (1 − γ^c_t) q_t + ρ_t n|_{A∖{a*}}` with mark `v(x) = Σ(p^c_t)^{−1} x` is
feasible in (3.6)", with `ρ_t = 16 B L_t² / t` (Definition G.2).

Explicit-hypothesis form: the properties of the event `W_t` used in the proof are assumed
directly — the reference is `a*`, `X_t ≥ 0` on `𝒦`, the ratio bound `Δ ≤ L X_t/t`, the two
feasibility thresholds in the form `ρ ≤ 1`, `ρ N_n ≤ min(Γ_t, 1/4)`, the atom mass
`q_t(a*) ≥ 3/4`, and the proxy budget `64 ∫_{a≠a*} e^{−X_t/(2β_t)} dµ ≤ 2d ρ² µ(a*)` supplied by
Lemma G.3. -/
theorem lemma_G_7 {d : ℕ} (hd : 1 ≤ d) (𝒜 ℬ : Set (Fin d → ℝ)) (e : ℕ → Fin d → ℚ)
    (hℬ : IsCompact ℬ) (hgen : IsGenerator 𝒜 ℬ) (hspan : Submodule.span ℝ 𝒜 = ⊤)
    (θ astar : Fin d → ℝ) (haK : astar ∈ Kset ℬ)
    (hΔ0 : ∀ a ∈ Kset ℬ, 0 ≤ (a - astar) ⬝ᵥ θ) (hΔ2 : ∀ a ∈ Kset ℬ, (a - astar) ⬝ᵥ θ ≤ 2)
    (ν : Measure (Fin d → ℝ)) (hν : IsCtrAlloc ℬ θ astar ν)
    (n : ℕ) (hn : 1 ≤ n) (S : Fin d → ℝ) (href : gref ℬ e n S = astar)
    (hX0 : ∀ a ∈ Kset ℬ, 0 ≤ (a - astar) ⬝ᵥ S) (L : ℝ) (hL : 1 ≤ L)
    (hgapL : ∀ a ∈ Kset ℬ, (a - astar) ⬝ᵥ θ ≤ L * ((a - astar) ⬝ᵥ S / n))
    (hρ1 : 16 * gB d * L ^ 2 / n ≤ 1)
    (hN : 16 * gB d * L ^ 2 / n * (ν Set.univ).toReal ≤ gGam d n)
    (hN4 : 16 * gB d * L ^ 2 / n * (ν Set.univ).toReal ≤ 1 / 4)
    (hq : 3 / 4 ≤ (gq ℬ e n S {astar}).toReal)
    (ha : 0 < refMu ℬ e {astar})
    (hprox : 64 * ∫ x in {astar}ᶜ, Real.exp (-((x - astar) ⬝ᵥ S) / (2 * gbeta d n))
        ∂(refMu ℬ e) ≤ 2 * d * (16 * gB d * L ^ 2 / n) ^ 2 * (refMu ℬ e {astar}).toReal) :
    IsFeasible3_6 ℬ e n S
      (ENNReal.ofReal (1 - ((ENNReal.ofReal (16 * gB d * L ^ 2 / n) •
          ν.restrict {astar}ᶜ) Set.univ).toReal) • gq ℬ e n S +
        ENNReal.ofReal (16 * gB d * L ^ 2 / n) • ν.restrict {astar}ᶜ) := by sorry

end IAFTRL
Source
Zhao–Tan, Best-of-Both-Worlds FTRL for Linear Bandits: The Interplay of Information Acquisition, Corruption, and Geometry (STOC submission version; arXiv:2611.01299), Appendix G.4.1, Definition G.2 and Lemma G.7, (G.114)–(G.140)

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