TY  - RPRT
TI  - Optimal Single-Policy Sample Complexity and Transient Coverage for Average-Reward Offline RL
AU  - Matthew Zurek
AU  - Guy Zamir
AU  - Yudong Chen
PY  - 2026
UR  - https://arxiv.org/abs/2506.20904
ID  - 2506.20904
ER  -