POMDP Planning on LightDark 10
17.5ReturnGammaZero
Evaluation Results
| Method | Links | |
|---|---|---|
| GammaZeroEvaluation Mode=Full2025.10 | 17.5 | |
| BetaZeroEvaluation Mode=Full2025.10 | 16.17 | |
| GammaZeroEvaluation Protocol=zero-shot transfer, Search Configuration=Full search2025.10 | 15.2 | |
| GammaZeroEvaluation Mode=Raw Pθ2025.10 | 14.4 | |
| BetaZeroEvaluation Mode=Raw Pθ2025.10 | 13.98 | |
| GammaZeroEvaluation Mode=Raw V*θ2025.10 | 13.3 | |
| BetaZeroEvaluation Mode=Raw V*θ2025.10 | 12.45 | |
| GammaZeroEvaluation Protocol=zero-shot transfer, Search Configuration=Raw policy network (Pθ)2025.10 | 12.1 | |
| GammaZeroEvaluation Protocol=zero-shot transfer, Search Configuration=One-step look-ahead value network (V*θ)2025.10 | 11.2 | |
| AdaOPSEvaluation Mode=Standard2025.10 | 6.28 | |
| AdaOPSEvaluation Protocol=per-size reference2025.10 | 6.28 | |
| POMCPOWEvaluation Mode=Standard2025.10 | 1.08 | |
| POMCPOWEvaluation Protocol=per-size reference2025.10 | 1.08 | |
| DESPOTEvaluation Mode=Standard2025.10 | 0.73 | |
| DESPOTEvaluation Protocol=per-size reference2025.10 | 0.73 |