Robot Manipulation on SimplerEnv WidowX Robot tasks
9,800Average Success RateIPI
Evaluation results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| IPIBase Model=OpenVLA-7B, Framework=StARe2025.12 | 9,800 | — | — | — | — | — | — | — | — | — | |
| StA-PPOBase Model=OpenVLA-7B, Fine-tuning Protocol=SFT → StA-PPO(StARe)2025.12 | 9,460 | — | — | — | — | — | — | — | — | — | |
| RL4VLABase Model=OpenVLA-7B, Fine-tuning Protocol=RL4VLA2025.12 | 9,250 | — | — | — | — | — | — | — | — | — | |
| IPIBase Model=Pi0.5, Framework=StARe2025.12 | 9,150 | — | — | — | — | — | — | — | — | — | |
| StA-PPOBase Model=Pi0.5, Fine-tuning Protocol=SFT → StA-PPO(StARe)2025.12 | 8,450 | — | — | — | — | — | — | — | — | — | |
| π_RLBase Model=Pi0.5, Fine-tuning Protocol=RL2025.12 | 7,960 | — | — | — | — | — | — | — | — | — | |
| StA-TPOBase Model=Pi0.5, Fine-tuning Protocol=SFT → StA-TPO(StARe)2025.12 | 6,050 | — | — | — | — | — | — | — | — | — | |
| SOFARBase Model=SOFAR2025.12 | 5,830 | — | — | — | — | — | — | — | — | — | |
| SFTBase Model=Pi0.5, Fine-tuning Protocol=SFT2025.12 | 5,710 | — | — | — | — | — | — | — | — | — | |
| GRAPEBase Model=Pi0.5, Fine-tuning Protocol=GRAPE2025.12 | 5,360 | — | — | — | — | — | — | — | — | — | |
| StA-TPOBase Model=OpenVLA-7B, Fine-tuning Protocol=SFT → StA-TPO(StARe)2025.12 | 5,160 | — | — | — | — | — | — | — | — | — | |
| GRAPEBase Model=OpenVLA-7B, Fine-tuning Protocol=GRAPE2025.12 | 4,390 | — | — | — | — | — | — | — | — | — | |
| SFTBase Model=OpenVLA-7B, Fine-tuning Protocol=SFT2025.12 | 4,170 | — | — | — | — | — | — | — | — | — | |
| SpatialVLABase Model=SpatialVLA2025.12 | 3,440 | — | — | — | — | — | — | — | — | — | |
| Octo-SmallBase Model=Octo-Small2025.12 | 3,000 | — | — | — | — | — | — | — | — | — | |
| Octo-BaseBase Model=Octo-Base2025.12 | 1,600 | — | — | — | — | — | — | — | — | — | |
| RoboVLMBase Model=RoboVLM2025.12 | 1,350 | — | — | — | — | — | — | — | — | — | |
| RT-1-XBase Model=RT-1-X2025.12 | 110 | — | — | — | — | — | — | — | — | — | |
| Green-VLAVariant=R2, Pre-training=Mixture-data including Bridge dataset2026.01 | 79.1 | 87.5 | 95.8 | 91.7 | 91.6 | 91.7 | 90.1 | 52.6 | 84.8 | 89 | |
| DB-MemVLAPre-training=Mixture-data including Bridge dataset, Runs=7, Episode Length=80 steps2026.01 | 73.2 | 91.7 | 83.3 | 79.2 | 100 | 88.6 | 85.1 | 57.6 | 100 | 50 | |
| FALCON2025.10 | 56.3 | — | — | — | — | — | 62.5 | 20.8 | 100 | 41.7 | |
| Green-VLAVariant=R1, Pre-training=Mixture-data including Bridge dataset2026.01 | 55.2 | 75 | 91.7 | 87.5 | 50 | 76.1 | 66.7 | 37.5 | 79.2 | 37.5 | |
| FlowerPre-training=Mixture-data including Bridge dataset2026.01 | 45 | — | — | — | — | — | 71 | 8 | 88 | 13 | |
| Green-VLAVariant=R0, Pre-training=Mixture-data including Bridge dataset2026.01 | 45 | 66.7 | 91.7 | 91.7 | 50 | 75 | 33.3 | 33.3 | 88.5 | 25 | |
| SpatialVLA2025.10 | 42.7 | — | — | — | — | — | 16.7 | 29.2 | 100 | 25 | |
| RoboVLM2025.10 | 37.5 | — | — | — | — | — | 45.8 | 4.2 | 79.2 | 20.8 | |
| π0Status=Fine-tuned, Pre-training=Mixture-data including Bridge dataset2026.01 | 27.1 | 45.8 | 50 | 91.6 | 25 | 53.1 | 29.1 | 16.7 | 62.5 | 0 | |
| Octo-Base2025.10 | 16 | — | — | — | — | — | 12.5 | 0 | 43.1 | 8.3 | |
| RT-1-XPre-training=Mixture-data including Bridge dataset2026.01 | 1.1 | 16.7 | 8.3 | 0 | 20.8 | 11.5 | 0 | 0 | 0 | 4.2 | |
| RT-1-X2025.10 | 1.1 | — | — | — | — | — | 0 | 0 | 0 | 4.2 | |
| OpenVLAPre-training=Mixture-data including Bridge dataset2026.01 | 1 | 4.1 | 12.5 | 8.3 | 33 | 14.5 | 0 | 0 | 4.1 | 0 | |
| OpenVLA2025.10 | 1 | — | — | — | — | — | 0 | 0 | 4.1 | 0 |