TY  - RPRT
TI  - Fast Last-Iterate Convergence in Zero-Sum Markov Games with Bandit Feedback
AU  - Yuheng Zhang
PY  - 2026
UR  - https://arxiv.org/abs/2610.05968
ID  - 2610.05968
ER  -