Visual Grounding on Single-Env (test)
88Success RateOneTwoVLA-VL
Evaluation results
| Method | Links | |
|---|---|---|
| OneTwoVLA-VLTraining Data=Robot data + 16,000 synthetic vision-language samples2025.05 | 88 | |
| OneTwoVLATraining Data=Exclusively robot data2025.05 | 78 | |
| π0Training Data=Exclusively robot data2025.05 | 5 |