Search papers, labs, and topics across Lattice.
Universit茅 Grenoble Alpes
1
0
2
4
Forget retraining: LEVER lets you snap together pre-trained RL policies at inference time, matching or beating from-scratch performance in some cases.