TY  - RPRT
TI  - Sharp Convergence and Sample Complexity of Policy Mirror Descent for Average-Reward MDPs
AU  - Enes Arda
AU  - Atilla Eryilmaz
PY  - 2026
UR  - https://arxiv.org/abs/2610.04117
ID  - 2610.04117
ER  -