Tag: offline RL algorithms
Browse our exclusive articles!
Android Coach: Boost Online Training with Multi-Action RL
Discover how Android Coach enhances online agent training efficiency by enabling multiple actions per state, improving success rates and reducing costs.
Advantage Reward Modeling for Long-Horizon Robotics
Discover ARM, a novel reward modeling method boosting long-horizon robotic manipulation with efficient labeling and improved reinforcement learning.
Offline Policy Optimization with Parametric Policies in RL
Explore advanced offline policy optimization using parametric policies to improve reinforcement learning in large and continuous action spaces.