Search papers, labs, and topics across Lattice.
Affiliation:
1
0
2
QWM achieves unprecedented sample efficiency in reinforcement learning by leveraging world models without succumbing to compounding bias, outperforming prior methods on challenging benchmarks.
