Robotic Manipulation on RLBench (test)
90.5Average Success RateTGM-VLA
Evaluation Results
| Method | Links | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| TGM-VLAReported Hardware=2A6000×1d2026.02 | 90.5 | 100 | 99 | 88 | 98 | 93 | 100 | 99 | 100 | 96 | 94 | 89 | 100 | 70 | 87 | 83 | 94 | 80 | 60 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TGM-VLA w/o ITM2026.02 | 88.8 | 100 | 99 | 92 | 99 | 96 | 98 | 98 | 100 | 95 | 92 | 86 | 100 | 65 | 82 | 82 | 91 | 78 | 50 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| BridgeVLAReported Hardware=48H100×20h2026.02 | 88.2 | 98.4 | 96 | 87.2 | 100 | 92.8 | 99.2 | 100 | 100 | 99.2 | 88 | 87.2 | 100 | 76.8 | 81.6 | 73.6 | 88 | 60.8 | 58.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TGM-VLA w/o CTM2026.02 | 88.2 | 100 | 97 | 98 | 96 | 96 | 95 | 99 | 98 | 94 | 94 | 88 | 100 | 62 | 79 | 80 | 92 | 73 | 47 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SAM2ACTReported Hardware=8H100×12h or 2A6000×5d2026.02 | 86.8 | 100 | 86 | 99 | 98 | 96 | 99 | 99 | 99 | 98 | 93 | 89 | 83 | 76 | 78 | 75 | 84 | 64 | 47 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FlowRAMNumber of demonstrations=1002025.06 | 84.9 | 100 | 100 | 92 | 94 | 100 | 92 | 96 | 100 | 96 | 96 | 84 | 92 | 77.3 | 61 | 86 | 72 | 48 | 42 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RVT-2Number of demonstrations=1002025.06 | 81.4 | 100 | 92 | 100 | 99 | 99 | 96 | 100 | 99 | 96 | 95 | 88 | 74 | 80 | 69 | 66 | 40 | 35 | 38 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RVT-2Train time=0.83 days, Inf. Speed=20.6 fps, Resolution=128x128, Number of demonstrations=1002024.06 | 81.4 | 100 | 92 | 100 | 99 | 99 | 96 | 100 | 99 | 96 | 95 | 88 | 74 | 80 | 69 | 66 | 40 | 35 | 38 | 1.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RVT-2Reported Hardware=8V100×20h2026.02 | 81.4 | 100 | 92 | 100 | 99 | 99 | 96 | 100 | 99 | 96 | 95 | 88 | 74 | 80 | 69 | 66 | 40 | 35 | 38 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| 3D Diffuser ActorNumber of demonstrations=1002025.06 | 81.3 | 98.4 | 97.6 | 84 | 96.8 | 99.2 | 96 | 96 | 100 | 97.6 | 93.6 | 82.4 | 89.6 | 68.3 | 47.2 | 85.6 | 65.6 | 44 | 24 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| 3D Diffuser ActorView setup=multi-view2024.02 | 81.3 | 98.4 | 97.6 | 84 | 96.8 | 99.2 | 96 | 96 | 100 | 97.6 | 93.6 | 82.4 | 89.6 | 68.3 | 47.2 | 85.6 | 65.6 | 44 | 24 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MLA2025.09 | 81 | — | — | 100 | — | — | — | — | — | — | 75 | — | — | — | — | — | — | — | — | — | — | — | 90 | — | — | — | — | — | — | 60 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 55 | — | — | — | 95 | — | 95 | — | 100 | 50 | — | — | 90 | 0.03 | |
| 3D Diffuser ActorAgent Type=3D next-best-pose, Priors=depth, keypoints, scene bounds, motion-planners, Training Demos=50, Protocol=multi-task2024.07 | 73.8 | 46.7 | 99.3 | — | — | 99.3 | — | — | — | — | — | — | 71.3 | — | — | — | — | — | — | — | 100 | 29.3 | 76 | 8 | 76.7 | 93.3 | 58 | 82.7 | 96.7 | 94 | 92.7 | 92 | 83.3 | 36.7 | 77.3 | 96 | 66 | 100 | 72.7 | 94 | 4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| HybridVLA2025.09 | 66 | — | — | 80 | — | — | — | — | — | — | 70 | — | — | — | — | — | — | — | — | — | — | — | 30 | — | — | — | — | — | — | 50 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 25 | — | — | — | 85 | — | 95 | — | 100 | 50 | — | — | 75 | 0.05 | |
| Act3DTrain time=5.0 days, Resolution=256x256, Number of demonstrations=1002024.06 | 65 | 99 | 93 | 92 | 94 | 94 | 90 | 92 | 92 | 95 | 80 | 47 | 93 | 12 | 9 | 51 | 27 | 8 | 3 | 2.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| π02025.09 | 65 | — | — | 85 | — | — | — | — | — | — | 65 | — | — | — | — | — | — | — | — | — | — | — | 90 | — | — | — | — | — | — | 5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 25 | — | — | — | 85 | — | 100 | — | 90 | 10 | — | — | 95 | 0.04 | |
| DreamVLA*status=re-implemented2025.09 | 65 | — | — | 25 | — | — | — | — | — | — | 85 | — | — | — | — | — | — | — | — | — | — | — | 50 | — | — | — | — | — | — | 35 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 35 | — | — | — | 95 | — | 100 | — | 95 | 55 | — | — | 75 | 0.05 | |
| FDPPolicy=FDP2025.12 | 63.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 72 | 100 | 75.5 | 93 | 27.5 | 78 | 61 | 4.5 | — | — | |
| Act3DView setup=multi-view2024.02 | 63.2 | 93.6 | 96 | 86.4 | 95.2 | 94.4 | 91.2 | 96.8 | 80.8 | 95.2 | 59.2 | 32.8 | 78.4 | 4 | 9.6 | 67.2 | 24 | 29.6 | 3.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RVTTrain time=1.0 days, Inf. Speed=11.6 fps, Resolution=128x128, Number of demonstrations=1002024.06 | 62.9 | 100 | 81.6 | 72 | 88 | 93.6 | 88 | 52 | 99.2 | 91.2 | 91 | 48 | 71.2 | 28.8 | 26.4 | 49.6 | 11.2 | 36 | 4 | 2.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RVTView setup=multi-view2024.02 | 62.9 | 100 | 81.6 | 72 | 88 | 93.6 | 88 | 52 | 99.2 | 91.2 | 91 | 48 | 71.2 | 28.8 | 26.4 | 49.6 | 11.2 | 36 | 4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RVTReported Hardware=8V100×1d2026.02 | 62.9 | 100 | 81.6 | 72 | 88 | 93.6 | 88 | 52 | 99.2 | 91.2 | 91 | 48 | 71.2 | 28.8 | 26.4 | 49.6 | 11.2 | 36 | 4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| GENIMAAgent Type=RGB-to-joint, Training Demos=50, Protocol=multi-task2024.07 | 49.6 | 76.7 | 12.7 | — | — | 71.3 | — | — | — | — | — | — | 77.3 | — | — | — | — | — | — | — | 50 | 26 | 94 | 79.3 | 85.3 | 48.7 | 46.7 | 46 | 69.3 | 18.7 | 36 | 90 | 72.7 | 14.7 | 12.7 | 26.7 | 11.3 | 48 | 21.3 | 93.3 | 12 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PerActNumber of demonstrations=1002025.06 | 49.4 | 92.8 | 74 | 52 | 70.4 | 88 | 51.2 | 55.2 | 89.6 | 86 | 44.8 | 17.6 | 88 | 26.4 | 2.4 | 28 | 5.6 | 16.8 | 2.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PerActTrain time=16.0 days, Inf. Speed=4.9 fps, Resolution=128x128, Number of demonstrations=1002024.06 | 49.4 | 92.8 | 74 | 52 | 70.4 | 88 | 51.2 | 55.2 | 89.6 | 84 | 44.8 | 17.6 | 88 | 26.4 | 2.4 | 28 | 5.6 | 16.8 | 2.4 | 4.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PerActView setup=multi-view2024.02 | 49.4 | 92.8 | 74 | 52 | 70.4 | 88 | 51.2 | 55.2 | 89.6 | 86 | 44.8 | 17.6 | 88 | 26.4 | 2.4 | 28 | 5.6 | 16.8 | 2.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PerActReported Hardware=8V100×16d2026.02 | 49.4 | 92.8 | 74 | 52 | 70.4 | 88 | 51.2 | 55.2 | 89.6 | 84 | 44.8 | 17.6 | 88 | 26.4 | 2.4 | 28 | 5.6 | 16.8 | 2.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PolarNetResolution=128x128, Number of demonstrations=1002024.06 | 46.4 | 96 | 56 | 52 | 100 | 80 | 32 | 36 | 92 | 84 | 40 | 44 | 84 | 4 | 8 | 12 | 4 | 12 | 0 | 4.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PolarNetView setup=multi-view2024.02 | 46 | 96 | 56 | 52 | 100 | 80 | 32 | 36 | 92 | 84 | 40 | 44 | 84 | 4 | 8 | 12 | 4 | 12 | 0 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SpatialVLA2025.09 | 46 | — | — | 20 | — | — | — | — | — | — | 15 | — | — | — | — | — | — | — | — | — | — | — | 40 | — | — | — | — | — | — | 15 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 30 | — | — | — | 80 | — | 80 | — | 85 | 25 | — | — | 70 | 0.03 | |
| DPPolicy=DP2025.12 | 45.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 33.5 | 100 | 77 | 65.5 | 64 | 19 | 3.5 | 2 | — | — | |
| HiveFormerResolution=128x128, Number of demonstrations=1002024.06 | 45.3 | 84 | 64 | 28 | 100 | 80 | 68 | 52 | 76 | 76 | 80 | 8 | 52 | 8 | 0 | 32 | 0 | 8 | 0 | 5.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| HiveFormerView setup=multi-view2024.02 | 45 | 84 | 64 | 28 | 100 | 80 | 68 | 52 | 76 | 76 | 80 | 8 | 52 | 8 | 0 | 32 | 0 | 8 | 0 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MoDEPolicy=MoDE2025.12 | 45 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 31 | 97 | 70.5 | 50.5 | 60 | 10.5 | 37 | 3.5 | — | — | |
| GEM3DBackbone=PerceiverIO, Pre-training=100K iterations, Policy Training=12K iterations, SE(3) augmentations=true2026.01 | 44.2 | — | — | — | 44.3 | 54 | — | 44.7 | — | 29.7 | 11.7 | — | 65.7 | 7.7 | — | — | — | — | — | 1 | — | — | — | — | — | — | — | — | — | 67.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 32.7 | 40 | 61 | 72 | — | — | — | — | — | — | — | — | — | — | |
| SDPPolicy=SDP2025.12 | 42.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 24 | 100 | 68 | 42.5 | 58.5 | 6.5 | 36 | 1 | — | — | |
| UP-VLA2025.09 | 42 | — | — | 10 | — | — | — | — | — | — | 20 | — | — | — | — | — | — | — | — | — | — | — | 55 | — | — | — | — | — | — | 15 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 20 | — | — | — | 80 | — | 80 | — | 65 | 35 | — | — | 40 | 0.04 | |
| OpenVLA2025.09 | 40 | — | — | 55 | — | — | — | — | — | — | 20 | — | — | — | — | — | — | — | — | — | — | — | 15 | — | — | — | — | — | — | 20 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 5 | — | — | — | 60 | — | 85 | — | 75 | 30 | — | — | 35 | 0.02 | |
| ACTAgent Type=RGB-to-joint, Training Demos=50, Protocol=multi-task2024.07 | 39.6 | 31.3 | 22 | — | — | 59.3 | — | — | — | — | — | — | 82.7 | — | — | — | — | — | — | — | 32.7 | 18 | 42 | 69.3 | 75.3 | 40 | 22.6 | 22.6 | 8 | 13.3 | 43.3 | 99.3 | 65.3 | 44 | 14.7 | 15.3 | 22.7 | 44.7 | 37.3 | 45.3 | 19.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ManiGaussian (w Distill)Backbone=PerceiverIO, Distillation=true2026.01 | 33.3 | — | — | — | 35.3 | 42.7 | — | 31 | — | 20.3 | 5.3 | — | 54.3 | 5 | — | — | — | — | — | 3 | — | — | — | — | — | — | — | — | — | 64.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 16.3 | 33.7 | 48.7 | 43 | — | — | — | — | — | — | — | — | — | — | |
| ManiGaussianBackbone=PerceiverIO, Representation=3D Gaussian Splatting2026.01 | 31.5 | — | — | — | 30 | 44 | — | 27.3 | — | 5.7 | 8 | — | 57.3 | 2 | — | — | — | — | — | 3.08 | — | — | — | — | — | — | — | — | — | 66.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 19.3 | 38.3 | 30.7 | 49.3 | — | — | — | — | — | — | — | — | — | — | |
| GNFactorBackbone=PerceiverIO, Representation=NeRF-based2026.01 | 29.2 | — | — | — | 32 | 40.7 | — | 32 | — | 11 | 5.7 | — | 51.7 | 3.7 | — | — | — | — | — | 3.33 | — | — | — | — | — | — | — | — | — | 62 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 22 | 35.7 | 35 | 19 | — | — | — | — | — | — | — | — | — | — | |
| PerActBackbone=PerceiverIO, Representation=Voxel-based 3D2026.01 | 21.9 | — | — | — | 28.3 | 43.7 | — | 19 | — | 10.7 | 3 | — | 38.7 | 4 | — | — | — | — | — | 4.58 | — | — | — | — | — | — | — | — | — | 51.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 8.3 | 26.7 | 16 | 13.7 | — | — | — | — | — | — | — | — | — | — | |
| SuSIEAgent Type=RGB-to-joint, Training Demos=50, Protocol=multi-task2024.07 | 21.8 | 7.3 | 0 | — | — | 32.7 | — | — | — | — | — | — | 67.3 | — | — | — | — | — | — | — | 5.3 | 0 | 21.3 | 36 | 26.6 | 26 | 10 | 2 | 24.6 | 1 | 24.6 | 40 | 74.7 | 1 | 4 | 6.7 | 1 | 72 | 4 | 50 | 6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| C2F-ARM-BCResolution=128x128, Number of demonstrations=1002024.06 | 20.1 | 72 | 16 | 0 | 20 | 68 | 4 | 24 | 24 | 12 | 8 | 8 | 20 | 0 | 0 | 0 | 4 | 8 | 0 | 5.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| C2F-ARM-BCView setup=multi-view2024.02 | 20.1 | 72 | 16 | 0 | 20 | 68 | 4 | 24 | 24 | 12 | 8 | 8 | 20 | 0 | 0 | 0 | 4 | 8 | 0 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DiffusionPolicyAgent Type=RGB-to-joint, Training Demos=50, Protocol=multi-task2024.07 | 2.3 | 2.7 | 0 | — | — | 20.7 | — | — | — | — | — | — | 0 | — | — | — | — | — | — | — | 0 | 0 | 0 | 3.3 | 6.7 | 0 | 0 | 2 | 0 | 2 | 0.7 | 1.3 | 22.7 | 0 | 2 | 0 | 0 | 1.3 | 0 | 2 | 4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Image-BC (CNN)Resolution=128x128, Number of demonstrations=1002024.06 | 1.3 | 0 | 0 | 0 | 0 | 8 | 8 | 0 | 0 | 4 | 0 | 0 | 4 | 0 | 0 | 0 | 0 | 0 | 0 | 7.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Image-BC (ViT)Resolution=128x128, Number of demonstrations=1002024.06 | 1.3 | 0 | 0 | 0 | 0 | 16 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 7.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Claude-sonnet-4-52026.02 | — | 51 | — | — | — | — | — | — | — | — | — | — | — | — | 63.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 30 | 36.7 | 20 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DeepSeek-V32026.02 | — | 70 | — | — | — | — | — | — | — | — | — | — | — | — | 64.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 90 | 42 | 32.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Gemini-2.5-pro2026.02 | — | 85.6 | — | — | — | — | — | — | — | — | — | — | — | — | 51.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 95 | 46.7 | 30 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| GPT-52026.02 | — | 58.2 | — | — | — | — | — | — | — | — | — | — | — | — | 45 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 80 | 33.3 | 15 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| GSRModel size=8B2026.02 | — | 82 | — | — | — | — | — | — | — | — | — | — | — | — | 68.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 92.5 | 52.5 | 41.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PERACTNumber of demonstrations=10, Number of evaluation episodes=25, Random seeds=3, Input modality=RGB-D2024.02 | — | 20 | 32 | 72 | 72 | 32 | 36 | 12 | — | 68 | 16 | 28 | — | — | 56 | 16 | 0 | — | 4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Qwen3Model size=8B2026.02 | — | 51 | — | — | — | — | — | — | — | — | — | — | — | — | 56.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 25 | 33.3 | 26.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RVTNumber of demonstrations=10, Number of evaluation episodes=25, Random seeds=3, Input modality=RGB-D2024.02 | — | 45.33 | 54.67 | 76 | 96 | 22.67 | 97.33 | 5.33 | 2.67 | 69.33 | 91.33 | 28 | — | 0 | 65.33 | 38.67 | 5.33 | — | 2.67 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VPDDNumber of demonstrations=10, Number of evaluation episodes=25, Random seeds=3, Input modality=RGB2024.02 | — | 60 | 70.67 | 40 | 88.67 | 37.33 | 66.67 | 56 | 73.33 | 73.33 | 24 | 61.33 | — | 56 | 58.67 | 70.67 | 30.67 | — | 30.67 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — |