Language-conditioned long-horizon robotic manipulation on CALVIN ABC→D
99.6Success Rate (1 Task)AVA-VLA
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| AVA-VLAProtocol=Zero-shot generalization, Setting=ABC→D2025.11 | 99.6 | 97.6 | 94.1 | 89.9 | 84.1 | 4.65 | |
| FLOWERProtocol=Zero-shot generalization, Setting=ABC→D2025.11 | 99.4 | 95.8 | 90.7 | 84.9 | 77.8 | 4.53 | |
| VLA-AdapterProtocol=Zero-shot generalization, Setting=ABC→D2025.11 | 99.1 | 94.6 | 88.8 | 82.8 | 76.5 | 4.42 | |
| UnifiedVLAProtocol=Zero-shot generalization, Setting=ABC→D2025.11 | 98.9 | 94.8 | 89 | 82.8 | 75.1 | 4.41 | |
| OpenVLA-OFTProtocol=Zero-shot generalization, Setting=ABC→D2025.11 | 96.9 | 92 | 85.7 | 80.4 | 72.9 | 4.28 | |
| MG-SelectTask=ABC → D, Zero-shot evaluation=true, Joint imitation learning=true, Test-time scaling framework=true2025.10 | 96.9 | 88 | 77.8 | 67.6 | 55.8 | 3.86 | |
| AffordanceVLA (full)Protocol=Zero-shot OOD, Rollouts=1000, Stage II=Enabled2026.06 | 96.8 | 92 | 87.5 | 80.8 | 75.9 | 4.33 | |
| SeerProtocol=Zero-shot generalization, Setting=ABC→D2025.11 | 96.3 | 91.6 | 86.1 | 80.3 | 74 | 4.28 | |
| Seer-LargeProtocol=Zero-shot OOD, Rollouts=10002026.06 | 96.3 | 91.6 | 86.1 | 80.3 | 74 | 4.28 | |
| CLOVERProtocol=Zero-shot OOD, Rollouts=10002026.06 | 96 | 83.5 | 70.8 | 57.5 | 45.4 | 3.53 | |
| π0-FASTTask=ABC → D, Zero-shot evaluation=true, reproduced=true2025.10 | 96 | 85.8 | 74.4 | 62.4 | 50.6 | 3.69 | |
| UniVLAProtocol=Zero-shot generalization, Setting=ABC→D2025.11 | 95.5 | 85.8 | 75.4 | 66.9 | 56.5 | 3.8 | |
| UniVLAProtocol=Zero-shot OOD, Rollouts=10002026.06 | 95.5 | 85.8 | 75.4 | 66.9 | 56.5 | 3.8 | |
| VPPProtocol=Zero-shot OOD, Rollouts=10002026.06 | 95.3 | 88.2 | 80.3 | 72.9 | 64.5 | 4.01 | |
| SeerProtocol=Zero-shot OOD, Rollouts=10002026.06 | 94.4 | 87.2 | 79.9 | 72.2 | 64.3 | 3.98 | |
| Pi0Protocol=Zero-shot OOD, Rollouts=10002026.06 | 93.8 | 85 | 76.7 | 68.6 | 60.1 | 3.84 | |
| AffordanceVLA (w/o stage II)Protocol=Zero-shot OOD, Rollouts=1000, Stage II=Disabled2026.06 | 93.4 | 84.7 | 75.4 | 68.1 | 58.9 | 3.81 | |
| OpenVLAProtocol=Zero-shot generalization, Setting=ABC→D2025.11 | 91.3 | 77.8 | 62 | 52.1 | 43.5 | 3.27 | |
| OpenVLAProtocol=Zero-shot OOD, Rollouts=10002026.06 | 91.3 | 77.8 | 62 | 52.1 | 43.5 | 3.27 | |
| SuSIEProtocol=Zero-shot OOD, Rollouts=10002026.06 | 87 | 69 | 49 | 38 | 26 | 2.69 | |
| GR-1Protocol=Zero-shot OOD, Rollouts=10002026.06 | 85.4 | 71.2 | 59.6 | 49.7 | 40.1 | 3.06 | |
| RoboFlamingoProtocol=Zero-shot OOD, Rollouts=10002026.06 | 82.4 | 61.9 | 46.6 | 33.1 | 23.5 | 2.48 |