Antmaze
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Offline Reinforcement Learning | Antmaze Medium play offline (target domain) | Target Domain Score (Normalized)398.8 | 42 | |
| Offline Reinforcement Learning | antmaze medium-play | Score84.8 | 35 | |
| Offline Reinforcement Learning | Antmaze umaze | Average Return96.7 | 24 | |
| Offline Reinforcement Learning | antmaze large-play | Score78.2 | 18 | |
| Offline Reinforcement Learning | antmaze medium-diverse | Score85 | 18 | |
| Navigation | AntMaze | Success Rate9,110 | 16 | |
| Navigation | AntMaze Small | Success Rate9,510 | 16 | |
| Offline Reinforcement Learning | Antmaze umaze-diverse | Average Return90.7 | 15 | |
| Offline Reinforcement Learning | antmaze medium-play v0 | Avg Normalized Score8,830 | 14 | |
| Offline Reinforcement Learning | antmaze umaze-diverse v0 | Avg Normalized Score88.5 | 14 | |
| Offline Reinforcement Learning | antmaze umaze v0 | Averaged Normalized Score98.6 | 14 | |
| Offline Reinforcement Learning | AntMaze large-diverse (l-d) | Normalized Score77.3 | 11 | |
| Offline Reinforcement Learning | AntMaze (large-play (l-p)) | Normalized Score70 | 11 | |
| Offline Reinforcement Learning | AntMaze medium-diverse (m-d) | Normalized Score79.5 | 11 | |
| Offline Reinforcement Learning | AntMaze umaze-diverse (u-d) | Normalized Score84 | 11 | |
| Offline Reinforcement Learning | AntMaze umaze | Normalized Score92.7 | 11 | |
| Offline Reinforcement Learning (Navigation) | AntMaze umaze-diverse D4RL (ud) | Expert Normalized Return94 | 10 | |
| Offline Reinforcement Learning (Navigation) | AntMaze umaze D4RL | Expert Normalized Return91 | 10 | |
| Goal Reaching | AntMaze large play v2 | Success Rate60 | 10 | |
| Goal Reaching | AntMaze medium play v2 | Success Rate80.6 | 10 | |
| Offline Reinforcement Learning | AntMaze Ultra-Diverse | Avg Normalized Score5,460 | 10 | |
| Offline Reinforcement Learning | AntMaze-Ultra-Play | Avg Normalized Score56.6 | 10 | |
| Offline Goal-conditioned Reinforcement Learning | antmaze large navigate oraclerep v0 | Task 1 Score87 | 9 | |
| Reinforcement Learning | AntMaze large-play D4RL | Average Episodic Return533 | 8 | |
| Reinforcement Learning | AntMaze large-diverse D4RL | Average Episodic Return493 | 8 |
Showing 25 of 89 rows