Robotic Manipulation on SimplerEnv WidowX v1 (test)
87.5Spoon Success RateStructVLA
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| StructVLA2026.03 | 87.5 | 75 | 45.8 | 91.7 | 75 | — | |
| UniVLA2026.03 | 83.3 | 66.7 | 33.3 | 95.8 | 69.8 | — | |
| VideoVLA2026.03 | 75 | 20.8 | 45.8 | 70.8 | 53.1 | — | |
| π0.5 + SG-LPState Grounding=True, Latent Prompt Training=True, Initialization=pretrained π0.52026.06 | 72.5 | 69.5 | — | 70.5 | 63.5 | 41.5 | |
| π0.5 + SG-LP + TTTState Grounding=True, Latent Prompt Training=True, Test-Time Training=True, Initialization=pretrained π0.52026.06 | 71 | 74.5 | — | 76 | 67.4 | 48 | |
| CogACT2026.06 | 58.3 | 37.5 | — | 91.7 | 52.1 | 20.8 | |
| ThinkAct2026.06 | 58.3 | 37.5 | — | 70.8 | 43.8 | 8.7 | |
| Octo-SmallScale=Small2026.03 | 47.2 | 9.7 | 4.2 | 56.9 | 29.5 | — | |
| Octo-Small2026.06 | 47.2 | 9.7 | — | 56.9 | 29.5 | 4.2 | |
| RoboVLMs2026.03 | 45.8 | 20.8 | 4.2 | 79.2 | 37.5 | — | |
| RoboVLM2026.06 | 45.8 | 20.8 | — | 79.2 | 37.5 | 4.2 | |
| π02026.06 | 45.8 | 36.1 | — | 81.9 | 47.6 | 26.4 | |
| InstructVLA2026.06 | 43.1 | 40.3 | — | 94.4 | 46.9 | 9.7 | |
| mimic-video2026.03 | 41.7 | 54.2 | 29.2 | 100 | 56.3 | — | |
| Magma2026.06 | 37.5 | 29.2 | — | 91.7 | 44.8 | 20.8 | |
| π0.5Description=Internal baseline2026.06 | 33.3 | 79.2 | — | 75 | 51.1 | 16.7 | |
| SpatialVLA2026.06 | 20.8 | 20.8 | — | 70.8 | 34.4 | 25 | |
| SpatialVLA2026.03 | 16.7 | 25 | 29.2 | 100 | 42.7 | — | |
| Octo-BaseScale=Base2026.03 | 12.5 | 31.9 | 0 | 43.1 | 16 | — | |
| Octo-Base2026.06 | 12.5 | 8.3 | — | 43.1 | 16 | 0 | |
| RT-1-X2026.03 | 0 | 4.2 | 0 | 0 | 1.1 | — | |
| OpenVLA2026.03 | 0 | 0 | 0 | 4.1 | 1 | — | |
| RT-1-X2026.06 | 0 | 4.2 | — | 0 | 1.1 | 0 | |
| OpenVLA2026.06 | 0 | 0 | — | 4.1 | 1 | 0 |