Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
1
0
1
By leveraging a pretrained RL policy as a sampling prior, this method achieves real-time adaptability in humanoid control without retraining, enhancing precision in complex tasks.