Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

p. 5, the claim before Corollary 12 — optimistic FTRL with Mᵗ = uᵗ⁻¹ and utilities in [0,1] is stable with κ = 2η

Proved
RegLearnGames.IndivRegret.oftrl_stability

by mikedeng1 · Oct 9, 2026 · Mathlib 0df444a (Lean v4.33.1)

no-regret-dynamicsoptimistic-ftrlp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1stability

Let R\mathcal RR be a regularizer on Rd\mathbb R^dRd that is 1-strongly convex with respect to ∥⋅∥1\|\cdot\|_1∥⋅∥1​ on the simplex Δ\DeltaΔ, and let η>0\eta>0η>0. Let u0,u1,…u^0,u^1,\dotsu0,u1,… be a utility sequence with every entry in [0,1][0,1][0,1], and let w0,w1,…w^0,w^1,\dotsw0,w1,… be a run of optimistic FTRL with regularizer R\mathcal RR, step size η\etaη and one-step recency bias Mt=ut−1M^t=u^{t-1}Mt=ut−1. Then for every t≥0t\ge0t≥0,

∥wt−wt+1∥1≤2η.\|w^t-w^{t+1}\|_1\le 2\eta .∥wt−wt+1∥1​≤2η.

This is the stability property with κ=2η\kappa=2\etaκ=2η that the paper invokes, citing Lemma 20, to apply Theorem 11 to optimistic FTRL; it is the step that turns the regret bound of Theorem 11 into the T1/4T^{1/4}T1/4 rate of Corollary 12.

Formalization Note The bound uxt∈[0,1]u^t_x\in[0,1]uxt​∈[0,1] is the paper's standing assumption ui:S→[0,1]u_i : S\to[0,1]ui​:S→[0,1] (§2), which makes every utility vector lie in [0,1]d[0,1]^d[0,1]d; it is what bounds ∥Mt−ut∥∞\|M^t-u^t\|_\infty∥Mt−ut∥∞​ and ∥Mt+1∥∞\|M^{t+1}\|_\infty∥Mt+1∥∞​ by 111. At T=1T=1T=1 the predictor is M1=u0M^1=u^0M1=u0, so the bound at t=0t=0t=0 is about the first move from w0=argmin⁡ΔRw^0=\operatorname{argmin}_\Delta\mathcal Rw0=argminΔ​R. The norm is ∥⋅∥1\|\cdot\|_1∥⋅∥1​.

Preamble
import Mathlib
import Definitions.Def_RegLearnGames_TotalRegret_Setting
Formal statement
namespace RegLearnGames.IndivRegret

theorem oftrl_stability {d : ℕ} (𝓡 : (Fin d → ℝ) → ℝ) (η : ℝ) (hη : 0 < η)
    (h𝓡 : RegLearnGames.TotalRegret.IsOneStronglyConvexL1 𝓡) (u w : ℕ → Fin d → ℝ)
    (hu : ∀ t x, u t x ∈ Set.Icc (0 : ℝ) 1)
    (hw : RegLearnGames.TotalRegret.IsOFTRL 𝓡 η (fun T => u (T - 1)) u w) :
    ∀ t, RegLearnGames.TotalRegret.l1 (w t - w (t + 1)) ≤ 2 * η := by sorry

end RegLearnGames.IndivRegret
Source
Syrgkanis, Agarwal, Luo, Schapire, Fast Convergence of Regularized Learning in Games, arXiv:1507.00407v5, p. 5, §3.2, the sentence before Corollary 12 ("stability property with κ = 2η (see Lemma 20 in the appendix)")
Human review
  • Endorsed by Shuze Chen · Oct 9, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 9, 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