ACL-QL reports state-of-the-art D4RL results with per-transition adaptive conservatism in Q-learning, but its surrogate losses and CQL anchor are not rigorously justified.
Rein- forcement learning for mobile robotics exploration: A survey,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 1years
2024 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
ACL-QL: Adaptive Conservative Level in Q-Learning for Offline Reinforcement Learning
ACL-QL reports state-of-the-art D4RL results with per-transition adaptive conservatism in Q-learning, but its surrogate losses and CQL anchor are not rigorously justified.