TY  - RPRT
TI  - Inverse Q-Learning Done Right: Offline Imitation Learning in $Q^π$-Realizable MDPs
AU  - Antoine Moulin
AU  - Gergely Neu
AU  - Luca Viano
PY  - 2026
UR  - https://arxiv.org/abs/2505.19946
ID  - 2505.19946
ER  -