BenchmarkPAW-GEN-10
Environmente07Place loaded boxes in container (bimanual)
- Episodes
- Rho 30 DiT-Flow 30 GR00T N1.7 30 MolmoAct 2 30 Pi 0.5 30
- Total duration
- Rho 20:32 DiT-Flow 17:37 GR00T N1.7 22:31 MolmoAct 2 20:17 Pi 0.5 19:39
- Mean duration
- Rho 41.1 s DiT-Flow 35.2 s GR00T N1.7 45.0 s MolmoAct 2 40.6 s Pi 0.5 39.3 s
- Success
- Rho 67% DiT-Flow 30% GR00T N1.7 27% MolmoAct 2 70% Pi 0.5 13%
- Manipulation density
- Rho Share of time manipulating, five definitionsJoints moving99.9%End-effector moving99.7%Gripper engaged65.3%Contact force > 5 N71.1%Gripper engaged or in contact82.0% DiT-Flow Share of time manipulating, five definitionsJoints moving99.1%End-effector moving88.7%Gripper engaged71.5%Contact force > 5 N69.6%Gripper engaged or in contact90.2% GR00T N1.7 Share of time manipulating, five definitionsJoints moving100.0%End-effector moving99.2%Gripper engaged65.2%Contact force > 5 N77.4%Gripper engaged or in contact89.2% MolmoAct 2 Share of time manipulating, five definitionsJoints moving100.0%End-effector moving99.6%Gripper engaged59.4%Contact force > 5 N77.6%Gripper engaged or in contact84.8% Pi 0.5 Share of time manipulating, five definitionsJoints moving100.0%End-effector moving99.7%Gripper engaged67.0%Contact force > 5 N81.5%Gripper engaged or in contact88.2%
- SPARC smoothness
- Rho Spectral arc length of the motionHow directly the arm moved, read from the shape of its speed spectrum. Unitless measure ≤ -1; the closer to -1, the smoother.The measure ignores how large a movement is, so a large and a small one compare directly. Shown here as the mean over this environment's measured episodes.Cutoff10 HzAmplitude threshold5%Zero-padding16×Signalend-effector translationBalasubramanian et al., On the analysis of movement smoothness, J Neuroeng Rehabil 2015. DiT-Flow Spectral arc length of the motionHow directly the arm moved, read from the shape of its speed spectrum. Unitless measure ≤ -1; the closer to -1, the smoother.The measure ignores how large a movement is, so a large and a small one compare directly. Shown here as the mean over this environment's measured episodes.Cutoff10 HzAmplitude threshold5%Zero-padding16×Signalend-effector translationBalasubramanian et al., On the analysis of movement smoothness, J Neuroeng Rehabil 2015. GR00T N1.7 Spectral arc length of the motionHow directly the arm moved, read from the shape of its speed spectrum. Unitless measure ≤ -1; the closer to -1, the smoother.The measure ignores how large a movement is, so a large and a small one compare directly. Shown here as the mean over this environment's measured episodes.Cutoff10 HzAmplitude threshold5%Zero-padding16×Signalend-effector translationBalasubramanian et al., On the analysis of movement smoothness, J Neuroeng Rehabil 2015. MolmoAct 2 Spectral arc length of the motionHow directly the arm moved, read from the shape of its speed spectrum. Unitless measure ≤ -1; the closer to -1, the smoother.The measure ignores how large a movement is, so a large and a small one compare directly. Shown here as the mean over this environment's measured episodes.Cutoff10 HzAmplitude threshold5%Zero-padding16×Signalend-effector translationBalasubramanian et al., On the analysis of movement smoothness, J Neuroeng Rehabil 2015. Pi 0.5 Spectral arc length of the motionHow directly the arm moved, read from the shape of its speed spectrum. Unitless measure ≤ -1; the closer to -1, the smoother.The measure ignores how large a movement is, so a large and a small one compare directly. Shown here as the mean over this environment's measured episodes.Cutoff10 HzAmplitude threshold5%Zero-padding16×Signalend-effector translationBalasubramanian et al., On the analysis of movement smoothness, J Neuroeng Rehabil 2015.
Episode
Held-out benchmark environment
0:000:39