List your company

Maze Navigation on Maze MCR variant (100 held-out trajectories)

0.057D_policyPPO
0.038080.165790.29350.42121Jun 20, 2025

Evaluation results

MethodLinks
PPOReference Policy=Optimal2025.060.057
MPPO BReference Policy=Optimal2025.060.076
MPPO AReference Policy=Optimal2025.060.084
MPPO AReference Policy=Bot A2025.060.471
MPPO BReference Policy=Bot B2025.060.481
PPOReference Policy=Bot B2025.060.492
PPOReference Policy=Bot A2025.060.509
MPPO BReference Policy=Bot A2025.060.521
MPPO AReference Policy=Bot B2025.060.53