Search papers, labs, and topics across Lattice.
Universitat Pompeu Fabra
1
0
2
CG-CMARL scales to multi-agent scenarios without the exponential blowup in complexity, outperforming fixed reward-shaping methods in achieving optimal coordination.