Navigation on GOAT-BENCH Strict Physical Regime full 2,669-subtask (val_unseen)
55.6Subtask SRAnyGoal
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| AnyGoalN=3, Details=Zero-shot, decentralized2026.06 | 55.6 | 11.8 | |
| AnyGoalN=2, Details=Zero-shot, decentralized2026.06 | 52.4 | 12.7 | |
| AnyGoalN=1, Details=Zero-shot, single agent2026.06 | 41.9 | 14.4 | |
| TANGODetails=Zero-shot, single-agent2026.06 | 32.1 | 16.5 | |
| SenseAct-NN Skill ChainDetails=End-to-end RL2026.06 | 29.5 | 11.3 | |
| Modular GOATDetails=Modular, zero-shot2026.06 | 24.9 | 17.2 |