Search papers, labs, and topics across Lattice.
1
0
2
WPO, a promising RL algorithm for continuous control, is now proven to converge linearly, finally putting it on solid theoretical footing.