| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| D4RL AntMaze navigation tasks v2 | GTP-BC | AntMaze U Success Rate84.2 | 10 | 5d ago | |
| D4RL Gym locomotion tasks v2 | GTP-BC | Return (HalfCheetah-m)48.6 | 10 | 5d ago | |
| DeepMind Control (DMC) suite seen/unseen embodiments | Meta-Controller | Hopper Hop Score49.1 | 9 | 3mo ago | |
| RoboMimic State-Based (proficient-human (ph) mixed-human (mh)) | DiffusionPolicy-C | Lift Success Rate (ph)100 | 5 | 1mo ago | |
| DeepMind Control suite | Meta-Controller | Hopper: Hop (Unseen Emb.)34.8 | 5 | 3mo ago | |
| Starpilot (10k) | FLAM | Mean Episodic Return9.1 | 3 | 3mo ago | |
| Starpilot 1k | Mean Episodic Return1.9 | 3 | 3mo ago | ||
| Bigfish 10k | FLAM | Mean Episodic Return8.8 | 3 | 3mo ago | |
| Bigfish (1k) | FLAM | Mean Episodic Return1.8 | 3 | 3mo ago | |
| Continuous control tasks | Meta-Controller | Inference Time (ms)17.8 | 2 | 3mo ago |
