Robotic Manipulation on Unseen Object
75Unscrew Cap Success RateLaST-HD (w/ unseen HD)
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| LaST-HD (w/ unseen HD)Training Setting=Trained with additional human data in unseen scenarios, Zero-shot Evaluation=false2026.06 | 75 | 60 | 70 | 60 | 45 | 40 | 58 | 56 | |
| LaST0 (w/ unseen HD)Training Setting=Trained with additional human data in unseen scenarios, Zero-shot Evaluation=false2026.06 | 65 | 50 | 65 | 50 | 35 | 35 | 49 | 46 | |
| Cosmos-PolicyTraining Setting=Trained on in-domain data only, Zero-shot Evaluation=true2026.06 | 60 | 10 | 60 | 10 | 20 | 5 | 28 | 26 | |
| LaST-HD (Mix-HD)Training Setting=Trained on in-domain data only, Zero-shot Evaluation=true2026.06 | 45 | 40 | 40 | 35 | 40 | 10 | 35 | 31 | |
| π0.5Training Setting=Trained on in-domain data only, Zero-shot Evaluation=true2026.06 | 40 | 40 | 50 | 60 | 15 | 10 | 36 | 30 | |
| LaST0Training Setting=Trained on in-domain data only, Zero-shot Evaluation=true2026.06 | 40 | 35 | 45 | 35 | 30 | 5 | 32 | 30 |