Standard model-based reinforcement learning (MBRL) approaches fit a
tran...
We present the largest and most comprehensive empirical study of pre-tra...
Offline goal-conditioned reinforcement learning (GCRL) promises
general-...
We propose State Matching Offline DIstribution Correction Estimation
(SM...
Reinforcement Learning (RL) agents in the real world must satisfy safety...
Ensuring safety for human-interactive robotics is important due to the
p...
Importance sampling-based estimators for off-policy evaluation (OPE) are...
Many reinforcement learning (RL) problems in practice are offline, learn...
For autonomous cars to drive safely and effectively, they must anticipat...