Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Signal laws coincide under a fiber-invisible symmetric perturbation

Proved
BanditAlgorithm.pmSignalMeasure_eq_symmetric_perturbation_of_fiber_sum

by Harry_Xu · Aug 5, 2026 · Mathlib 0df444a (Lean v4.33.1)

bandit-theoryinformation-theorypartial-monitoring

Let a finite partial-monitoring game have signal map Φ\PhiΦ. Fix an action ccc, a base outcome distribution uuu, a perturbation direction qqq, and a scale Δ\DeltaΔ. Assume both symmetric perturbations u−Δqu-\Delta qu−Δq and u+Δqu+\Delta qu+Δq are probability distributions. If the perturbation is invisible in every signal fiber of action ccc, meaning

∑i: Φ(c,i)=σqi=0for every signal σ,\sum_{i:\,\Phi(c,i)=\sigma}q_i=0 \qquad\text{for every signal }\sigma,i:Φ(c,i)=σ∑​qi​=0for every signal σ,

then action ccc induces exactly the same signal law under the two perturbed environments:

Sc(u−Δq)=Sc(u+Δq).S_c(u-\Delta q)=S_c(u+\Delta q).Sc​(u−Δq)=Sc​(u+Δq).

This is the finite signal-law identity used for actions in the neighborhood NabN_{ab}Nab​ in the hard-game lower bound for partial monitoring. It isolates the precise consequence of the kernel/fiber condition in equation (37.6).

Preamble
import Definitions.Def_PartialMonitoringStochastic
import Mathlib.Data.ENNReal.BigOperators

open MeasureTheory ProbabilityTheory
open scoped BigOperators ENNReal

namespace BanditAlgorithm
Formal statement
theorem pmSignalMeasure_eq_symmetric_perturbation_of_fiber_sum
    {k d : ℕ} {𝕊 : Type*} [Fintype 𝕊] [MeasurableSpace 𝕊]
    [DecidableEq 𝕊]
    (G : PartialMonitoringGame k d 𝕊) (c : Fin k)
    (u q : Fin d → ℝ) (Δ : ℝ)
    (hminus : (fun i => u i - Δ * q i) ∈ stdSimplex ℝ (Fin d))
    (hplus : (fun i => u i + Δ * q i) ∈ stdSimplex ℝ (Fin d))
    (hfiber : ∀ σ : 𝕊,
      ∑ i ∈ Finset.univ.filter (fun i => G.Φ c i = σ), q i = 0) :
    pmSignalMeasure G (fun i => u i - Δ * q i) c =
      pmSignalMeasure G (fun i => u i + Δ * q i) c := by sorry
Source
Lattimore and Szepesvári, Bandit Algorithms (2020), Theorem 37.12, Step 2, printed p. 490; equation (37.6), https://tor-lattimore.com/downloads/book/book.pdf

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