Navigation on FrozenLake
27Success RateActFocus
Evaluation Results
| Method | Links | |
|---|---|---|
| ActFocusBackbone=3B, Evaluation Step=2002026.05 | 27 | |
| ActFocusAlgorithm=PPO, Model Scale=3B2026.05 | 27 | |
| UniformAlgorithm=GRPO, Model Scale=3B2026.05 | 26.2 | |
| ActFocusAlgorithm=GRPO, Model Scale=0.5B2026.05 | 25.2 | |
| UniformAlgorithm=PPO, Model Scale=3B2026.05 | 24.6 | |
| UniformAlgorithm=GRPO, Model Scale=0.5B2026.05 | 23.8 | |
| ActFocusAlgorithm=PPO, Model Scale=0.5B2026.05 | 20.9 | |
| ActFocusAlgorithm=GRPO, Model Scale=3B2026.05 | 19.1 | |
| UniformBackbone=3B, Evaluation Step=2002026.05 | 16.6 | |
| UniformAlgorithm=PPO, Model Scale=0.5B2026.05 | 16.2 | |
| ActFocusBackbone=0.5B, Evaluation Step=2002026.05 | 15.6 | |
| UniformBackbone=0.5B, Evaluation Step=2002026.05 | 5.1 |