Robotic Manipulation on SIMPLER novel YCB and GSO objects 1.0 (test)
96Success Count: Green CubeVideoVLA
Evaluation Results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| VideoVLArobot=Google Robot, eval_mode=Zero-shot generalization, backbone=CogVideoX-5B2025.12 | 96 | 84 | 88 | 40 | 72 | 80 | 68 | 44 | 28 | 52 | 65.2 | |
| CogACTrobot=Google Robot, eval_mode=Zero-shot generalization2025.12 | 84 | 72 | 64 | 16 | 32 | 44 | 20 | 32 | 20 | 40 | 42.4 | |
| SpatialVLArobot=Google Robot, eval_mode=Zero-shot generalization2025.12 | 76 | 52 | 84 | 32 | 60 | 56 | 32 | 56 | 32 | 28 | 50.8 | |
| π0robot=Google Robot, eval_mode=Zero-shot generalization2025.12 | 48 | 60 | 44 | 16 | 12 | 44 | 24 | 16 | 0 | 24 | 28.8 | |
| OpenVLArobot=Google Robot, eval_mode=Zero-shot generalization2025.12 | 12 | 24 | 12 | 0 | 0 | 4 | 0 | 4 | 4 | 4 | 6.4 |