Search papers, labs, and topics across Lattice.
Affiliation:
1
0
2
11
Relying on a single simulator in multi-agent RL leads to dangerous mode collapse, but innovative solutions can boost generalization and performance by up to 14%.
