AFTD Auto-Formalizing Theoretical Domains

prefixPotential_eq_potential

theoremverifiedimported Lean

Prefix potential equals the induced Hedge potential. Let $G$ be a finite two-player zero-sum game with payoff matrix $A$ having $M$ rows and $N$ columns, fix a learning rate $\eta \in \bbr$, and let $a_0, \dots, a_{T-1}$ be a sequence of $T$ column actions. For any round $t < T$, the prefix potential of $G$ at learning rate $\eta$ formed from the first $t$ column actions $a_0, \dots, a_{t-1}$ equals the Hedge potential at time $t$ of the loss sequence $\ell_s(i) = 1 - A_{i, a_s}$ induced on the row player by $G$ and the full sequence of column actions; that is, \[ \sum_{i \in \mathrm{Fin}\,M} \exp\!\Bigl(-\eta \sum_{s=0}^{t-1}\bigl(1 - A_{i, a_s}\bigr)\Bigr) \;=\; \sum_{i \in \mathrm{Fin}\,M} \exp\!\bigl(-\eta\, L_t(i)\bigr), \] where $L_t(i) = \sum_{s < t} \ell_s(i)$ is the cumulative loss of row $i$.

Statement

set_option relaxedAutoImplicit false in
set_option autoImplicit false in
open Real Finset BigOperators in
lemma prefixPotential_eq_potential {M N T : ℕ} (G : ZeroSumGame M N)
    (η : ℝ) (actions : Fin T → Fin N) (t : Fin T) :
    prefixPotential G η
        (t := t.val) (fun s : Fin t.val => actions ⟨s.val, lt_trans s.isLt t.isLt⟩) =
      potential η (G.toLossSeq actions) t.val
Source
TCSlib, `prefixPotential_eq_potential`
Lean proof
By Karim Abdel Sadek, Mark Bedaywi, from github.com/Shilun-Allan-Li/tcslib at c8a02591b1e5 (Copyright (c) 2026 Karim Abdel Sadek and Mark Bedaywi. All rights reserved, Apache-2.0). Compiled here against leanprover/lean4:v4.33.1, Mathlib v4.33.1; written for leanprover/lean4:v4.25.0.
Verified
07 Oct 2026
Axioms
Classical.choiceQuot.soundpropext
Built on 5
  • prefixPotentialThe prefix potential is the sum of prefix Hedge weights over all rows: \[ \Phi_t \;=\; \sum_{i \in \mathrm{Fin}\,M} w_t(i). \]
  • ZeroSumGameA finite two-player zero-sum game with $M$ row actions and $N$ column actions is a record consisting of a payoff matrix…
  • prefixHedgeWeight_eq_hedgeWeightPrefix Hedge weight agrees with the induced-loss Hedge weight. Let $G$ be a finite two-player zero-sum game with $M$ row actions and $N$…
  • ZeroSumGame.toLossSeqGiven a game $G$ and a sequence of column responses $j_0, \dots, j_{T-1}$, the induced loss sequence for the row player is…
  • potentialThe potential (sum of unnormalized weights) at time $t$ is \[ W_t \;=\; \sum_{i=1}^{N} w_t(i). \] The potential telescopes to…
Used by 1

Lean source view module on GitHub

set_option relaxedAutoImplicit false inset_option autoImplicit false inopen Real Finset BigOperators in/-- The prefix potential over the first `t` actions equals `potential` at time `t` of the induced loss sequence. -/lemma prefixPotential_eq_potential {M N T : ℕ} (G : ZeroSumGame M N)    (η : ℝ) (actions : Fin T → Fin N) (t : Fin T) :    prefixPotential G η        (t := t.val) (fun s : Fin t.val => actions ⟨s.val, lt_trans s.isLt t.isLt⟩) =      potential η (G.toLossSeq actions) t.val := by  simp only [prefixPotential, potential]  apply Finset.sum_congr rfl  intro i _  exact prefixHedgeWeight_eq_hedgeWeight G η actions t i
See it in the graphDiscuss this resultChallenge it