Robotic Manipulation on RMBench
51Battery TryOurs(DP)
Evaluation Results
| Method | Links | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Ours(DP)Skill Executor=Pre-trained DP2026.06 | 51 | — | 90 | 92 | 72 | 85 | — | 29 | 69 | — | — | 70 | — | |
| Goal2SkillEvaluation Protocol=SR@1002026.04 | 46 | 8 | 38 | — | — | — | 23 | 60 | — | 10 | 38.7 | 32.4 | — | |
| WLA-0subtask prediction=enabled2026.06 | 45 | — | — | — | — | — | — | 23 | 84 | 74 | — | — | 56.5 | |
| Ours(CS)Skill Executor=Concept-Grounded Skills2026.06 | 45 | — | 95 | 86 | 74 | 73 | — | 26 | 64 | — | — | 66 | — | |
| -Llanglanguage-based subtask prediction loss=removed2026.06 | 38 | — | — | — | — | — | — | 12 | 18 | 1 | — | — | 17.3 | |
| CortexNumber of synthesized demonstrations=50, Number of rollouts=1002026.07 | 37 | 14 | 100 | 100 | 99 | 63 | — | — | — | 20 | — | — | — | |
| Mem-0Number of demonstrations=50, Number of rollouts=1002026.03 | 28 | 4 | 89 | 90 | 67 | 14 | 52.8 | 18 | 68 | 0 | 28.5 | 42 | — | |
| Mem-0Type=Memory-based2026.06 | 28 | — | — | — | — | — | — | 18 | 68 | 0 | — | — | 28.5 | |
| HiMem-WAM2026.06 | 28 | 28 | 33 | 32 | 38 | 27 | 31.6 | 24 | 19 | 8 | 19.8 | 26.3 | — | |
| Mem-0Manual implementation status=true2026.06 | 28 | — | 89 | 90 | 67 | 14 | — | 18 | 68 | — | — | 53 | — | |
| Mem-0Number of synthesized demonstrations=50, Number of rollouts=1002026.07 | 28 | 4 | 89 | 90 | 67 | 14 | — | — | — | 0 | — | — | — | |
| X-VLANumber of demonstrations=50, Number of rollouts=1002026.03 | 26 | 9 | 13 | 18 | 16 | 3 | 11.8 | 1 | 2 | 0 | 7.3 | 9.8 | — | |
| X-VLAEvaluation Protocol=SR@1002026.04 | 26 | 9 | 13 | — | — | — | 11 | 1 | — | 0 | 9 | 9.8 | — | |
| X-VLAArchitecture=VLA2026.06 | 26 | — | — | — | — | — | — | 1 | 2 | 0 | — | — | 7.3 | |
| X-VLA2026.06 | 26 | 9 | 13 | 18 | 16 | 3 | 11.8 | 1 | 0 | 2 | 7.3 | 9.8 | — | |
| X-VLANumber of synthesized demonstrations=50, Number of rollouts=1002026.07 | 26 | 9 | 13 | 18 | 16 | 3 | — | — | — | 0 | — | — | — | |
| ACTNumber of demonstrations=50, Number of rollouts=1002026.03 | 19 | 1 | 29 | 0 | 2 | 2 | 6.8 | 0 | 0 | 0 | 4.8 | 5.9 | — | |
| ACTEvaluation Protocol=SR@1002026.04 | 19 | 1 | 29 | — | — | — | 15 | 0 | — | 0 | 6.3 | 9.8 | — | |
| ACT2026.06 | 19 | 1 | 29 | 0 | 2 | 2 | 6.8 | 0 | 0 | 0 | 4.8 | 5.9 | — | |
| ACTNumber of synthesized demonstrations=50, Number of rollouts=1002026.07 | 19 | 1 | 29 | 0 | 2 | 2 | — | — | — | 0 | — | — | — | |
| Pi0.5Number of demonstrations=50, Number of rollouts=1002026.03 | 16 | 9 | 13 | 11 | 24 | 15 | 14.4 | 6 | 0 | 0 | 5.5 | 10.4 | — | |
| Pi0.5Evaluation Protocol=SR@1002026.04 | 16 | 9 | 13 | — | — | — | 11 | 6 | — | 0 | 7.3 | 8.8 | — | |
| π0.5Architecture=VLA2026.06 | 16 | — | — | — | — | — | — | 6 | 0 | 0 | — | — | 5.5 | |
| Fast-WAMArchitecture=WAM2026.06 | 16 | — | — | — | — | — | — | 37 | 0 | 0 | — | — | 13.3 | |
| π0.52026.06 | 16 | 9 | 13 | 11 | 24 | 15 | 14.4 | 6 | 2 | 1 | 6.3 | 10.8 | — | |
| π0.5Type=reactive finetuned VLA2026.06 | 16 | — | 13 | 11 | 24 | 15 | — | 6 | 0 | — | — | 12 | — | |
| π0.5Number of synthesized demonstrations=50, Number of rollouts=1002026.07 | 16 | 9 | 13 | 11 | 24 | 15 | — | — | — | 0 | — | — | — | |
| MemER2026.06 | 14 | — | 24 | 42 | 51 | 60 | — | 2 | 25 | — | — | 31 | — | |
| Mem-VLAManual implementation status=true2026.06 | 12 | — | 20 | 35 | 42 | 55 | — | 0 | 31 | — | — | 28 | — | |
| DPNumber of demonstrations=50, Number of rollouts=1002026.03 | 10 | 1 | 0 | 0 | 11 | 20 | 6.4 | 10 | 0 | 0 | 5 | 5.8 | — | |
| DPEvaluation Protocol=SR@1002026.04 | 10 | 1 | 0 | — | — | — | 0.5 | 10 | — | 0 | 6.7 | 4.2 | — | |
| DP2026.06 | 10 | 1 | 0 | 0 | 11 | 2 | 2.8 | 10 | 0 | 0 | 5 | 3.8 | — | |
| DPType=non-pretrained baseline2026.06 | 10 | — | 0 | 0 | 11 | 20 | — | 10 | 0 | — | — | 7 | — | |
| DPNumber of synthesized demonstrations=50, Number of rollouts=1002026.07 | 10 | 1 | 0 | 0 | 11 | 20 | — | — | — | 0 | — | — | — | |
| ACT2026.03 | — | 2 | 20 | 0 | — | 2 | — | — | — | — | — | 5.5 | — | |
| DP2026.03 | — | 2 | 0 | 0 | — | 16 | — | — | — | — | — | 4.5 | — | |
| MemoAct2026.03 | — | 4 | 98 | 41 | — | 55 | — | — | — | — | — | 49.5 | — | |
| MVMP2026.03 | — | 4 | 21 | 34 | — | 55 | — | — | — | — | — | 28.5 | — | |
| SAMP2026.03 | — | 3 | 47 | 23 | — | 26 | — | — | — | — | — | 24.75 | — |