Gym halfcheetah-medium-replay
1 papers with code • 1 benchmarks • 1 datasets
D4RL
Most implemented papers
Uncertainty-Based Offline Reinforcement Learning with Diversified Q-Ensemble
corl-team/CORL
•
•
NeurIPS 2021
However, prior methods typically require accurate estimation of the behavior policy or sampling from OOD data points, which themselves can be a non-trivial problem.
