LunarLander-v2
PulseAugur coverage of LunarLander-v2 — every cluster mentioning LunarLander-v2 across labs, papers, and developer communities, ranked by signal.
-
PPO training stabilized by dropping redundant state transitions
Researchers have developed a method to improve the stability of reinforcement learning training by randomly dropping a fraction of transitions from on-policy rollouts. This technique, applied to Proximal Policy Optimiza…
-
New architecture improves multi-timescale reinforcement learning
Researchers have developed a new architecture called Target Decoupling to address issues in multi-timescale reinforcement learning. This approach separates short-term and long-term signals to improve policy updates, pre…
-
New RL Architecture Solves Multi-Timescale Signal Pathologies
Researchers have identified algorithmic pathologies in multi-timescale reinforcement learning when combining short-term and long-term signals. They propose a Target Decoupling architecture that separates temporal predic…