D4RL AntMaze
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Offline Reinforcement Learning | D4RL AntMaze-Medium Diverse | Normalized Performance94.2 | 15 | |
| Navigation | D4RL AntMaze umaze v2 | Initial D4RL Score137.4 | 12 | |
| Offline Reinforcement Learning | D4RL AntMaze medium-diverse v2 (test) | Normalized Score80.2 | 12 | |
| Offline Reinforcement Learning | D4RL AntMaze umaze-diverse v2 (test) | Score (Normalized)84 | 12 | |
| Behavior Cloning | D4RL AntMaze navigation tasks v2 | AntMaze U Success Rate84.2 | 10 | |
| Navigation | D4RL AntMaze large-diverse v2 | Normalized D4RL Score58.4 | 10 | |
| Navigation | D4RL AntMaze large-play v2 | Normalized D4RL Score55.83 | 10 | |
| Navigation | D4RL AntMaze medium-diverse v2 | Normalized D4RL Score85.5 | 10 | |
| Navigation | D4RL AntMaze medium-play v2 | Normalized D4RL Score81 | 10 | |
| Offline Reinforcement Learning | D4RL AntMaze umaze, umaze-d, med-p, med-d, large-p, large-d v0 | umaze Return96.8 | 10 | |
| Offline Reinforcement Learning | D4RL AntMaze v0 (test) | Umaze Score87.5 | 10 | |
| Offline Reinforcement Learning | D4RL AntMaze fixed, play, diverse | AntMaze UMaze (Fixed) Score85 | 10 | |
| Offline Reinforcement Learning | D4RL Antmaze umaze, medium, large v0 | AntMaze UMaze v0 Score92.3 | 9 | |
| Navigation | D4RL AntMaze Total | Total Normalized Score397.6 | 8 | |
| Offline Reinforcement Learning | D4RL Antmaze umaze, medium, large v2 | UMaze Score0.967 | 8 | |
| Offline Reinforcement Learning | D4RL Antmaze-Diverse | AntMaze-Medium9,240 | 8 | |
| Offline Reinforcement Learning | D4RL Antmaze (Play) | AntMaze Medium Score88.8 | 8 | |
| Offline Reinforcement Learning | D4RL AntMaze Large-Diverse (AM-LD) | Normalized Score71 | 7 | |
| Offline multitask Reinforcement Learning | D4RL Antmaze large-play | Average Episodic Return306 | 7 | |
| Offline multitask Reinforcement Learning | D4RL Antmaze large-diverse | Avg Episodic Return359 | 7 | |
| Offline multitask Reinforcement Learning | D4RL Antmaze medium-play | Average Episodic Return624 | 7 | |
| Offline multitask Reinforcement Learning | D4RL Antmaze medium-diverse | Episodic Return631 | 7 | |
| Offline multitask Reinforcement Learning | D4RL Antmaze umaze-diverse | Average Episodic Return577 | 7 | |
| Offline multitask Reinforcement Learning | D4RL Antmaze umaze | Average Episodic Return593 | 7 | |
| Offline Goal-conditioned Reinforcement Learning | D4RL AntMaze large-diverse v2 | Normalized Score54.1 | 7 |
Showing 25 of 40 rows