Bosslevel on BabyAI
0.343Average Pass RatePolychromic PPO
Evaluation Results
| Method | Links | |
|---|---|---|
| Polychromic PPOevaluation_protocol=pass@1, initial-state perturbations=large2025.09 | 0.343 | |
| Poly-PPO w/ UCBevaluation_protocol=pass@1, initial-state perturbations=large2025.09 | 0.328 | |
| PPO w/ UCBevaluation_protocol=pass@1, initial-state perturbations=large2025.09 | 0.282 | |
| REINFORCE w/ UCBevaluation_protocol=pass@1, initial-state perturbations=large2025.09 | 0.276 | |
| PPOevaluation_protocol=pass@1, initial-state perturbations=large2025.09 | 0.266 | |
| Pretrained policyevaluation_protocol=pass@1, initial-state perturbations=large2025.09 | 0.238 | |
| REINFORCE w/ Baselineevaluation_protocol=pass@1, initial-state perturbations=large2025.09 | 0.225 |