Coffee Task on MimicGen
99.2Success RateOurs IL
Evaluation Results
| Method | Links | |
|---|---|---|
| Ours ILNumber of demos=10002025.09 | 99.2 | |
| Ours EnsNumber of demos=10002025.09 | 98.8 | |
| Ours EnsNumber of demos=2002025.09 | 96.4 | |
| Ours EnsNumber of demos=5002025.09 | 96.4 | |
| MimicGenNumber of demos=10002025.09 | 96.4 | |
| Ours ILNumber of demos=2002025.09 | 96 | |
| MimicGenNumber of demos=5002025.09 | 95.6 | |
| Ours ILNumber of demos=5002025.09 | 95.2 | |
| Ours EnsNumber of demos=502025.09 | 93.6 | |
| Ours EnsNumber of demos=1002025.09 | 93.2 | |
| MimicGenNumber of demos=1002025.09 | 92.4 | |
| MimicGenNumber of demos=2002025.09 | 92.4 | |
| Ours ILNumber of demos=1002025.09 | 89.6 | |
| LLM TrainerOptimization=Best Annotation2025.09 | 89 | |
| Ours ILNumber of demos=502025.09 | 86.4 | |
| MimicGenNumber of demos=502025.09 | 84 | |
| LLM TrainerOptimization=Total2025.09 | 82.6 | |
| MimicGenConfig=Baseline2025.09 | 78.2 | |
| LLM TrainerOptimization=No RL2025.09 | 39 |