Sharp Convergence and Sample Complexity of Policy Mirror Descent for Average-Reward MDPs 65E9N3DFYP

URI	`m2://res/65E9N3DFYP`
Kind	preprint
Creators	Enes Arda, Atilla Eryilmaz
Date	2026-10-02
Subjects	arxiv cs.LG, arxiv math.OC
Scope	border

Expressions

e1: version, preprint

Date	2026-10-02
Ids	arXiv 2610.04117

Items in this resource

The mentions of the inventory that name this resource. Each name links to the card of its inventory row.

Item	Kind	Where	Source
Poisson equation and differential values	statement	fact B.1	arxiv
Tabular critic upper bound, detailed form	statement	lemma C.2	arxiv
LFA critic lower bound	statement	proposition 5.9	arxiv
Finite mixing time for a fixed policy	statement	lemma E.1	arxiv
54FFZ00Q2B	statement	theorem	arxiv
Master convergence for inexact PMD	statement	lemma 3.7	arxiv
Tabular critic upper bound	statement	lemma 4.6	arxiv
LFA critic lower bound, detailed form	statement	proposition D.5	arxiv
Tabular sample complexity, detailed form	statement	theorem C.4	arxiv
Linear convergence of inexact PMD under LFA	statement	theorem D.1	arxiv
9HETP3AZQN	statement	lemma 3.2	arxiv
Uniform boundedness of stationary distributions	statement	lemma B.2	arxiv
LFA critic upper bound	statement	lemma D.3	arxiv
Two-state mixing-exponent lower bound, detailed form	statement	proposition C.5	arxiv
Policy gradient theorem	statement	fact 3.4	arxiv
FJFF27DB3D	definition	definition	arxiv
Tabular sample complexity	statement	theorem 4.8	arxiv
Linear convergence of inexact PMD under LFA	statement	theorem 5.6	arxiv
Linear convergence of tabular inexact PMD	statement	theorem C.1	arxiv
Exact PMD and superlinear convergence	statement	corollary 4.3	arxiv
Estimating $w_t^\star$ with an unbiased label oracle	statement	lemma D.2	arxiv
LFA sample complexity	statement	theorem 5.11	arxiv
Performance difference lemma	statement	fact 3.3	arxiv
LFA critic upper bound	statement	lemma 5.8	arxiv
Three-point inequality for PMD	statement	fact 3.6	arxiv
Tabular critic lower bound	statement	proposition 4.7	arxiv
Linear convergence of tabular inexact PMD	statement	theorem 4.2	arxiv
Covariance summability under uniform mixing	statement	lemma C.3	arxiv
LFA sample complexity	statement	theorem D.4	arxiv
Uniform mixing	statement	lemma 4.4	arxiv

Catalogue records

arxiv: 2610.04117 (work)