BenchmarkPAW-GEN-10
Environmente06Put tools in standing toolbox
- Episodes
- MolmoAct 2 30 Claude Fable 5.2 (D10) 30 Claude Opus 5.5 (D10) 30 DiT-Flow 30 Gemini 4 Argon (D10) 30 GPT-6 Astra (D10) 30 GR00T N1.7 30 Pi 0.5 30 Rho 30
- Total duration
- MolmoAct 2 13:51 Claude Fable 5.2 (D10) 105:07 Claude Opus 5.5 (D10) 51:22 DiT-Flow 8:15 Gemini 4 Argon (D10) 64:14 GPT-6 Astra (D10) 30:06 GR00T N1.7 12:29 Pi 0.5 13:43 Rho 12:54
- Mean duration
- MolmoAct 2 27.7 s Claude Fable 5.2 (D10) 210.2 s Claude Opus 5.5 (D10) 102.7 s DiT-Flow 16.5 s Gemini 4 Argon (D10) 128.5 s GPT-6 Astra (D10) 60.2 s GR00T N1.7 25.0 s Pi 0.5 27.4 s Rho 25.8 s
- Success
- MolmoAct 2 3% Claude Fable 5.2 (D10) 97% Claude Opus 5.5 (D10) 7% DiT-Flow 0% Gemini 4 Argon (D10) 0% GPT-6 Astra (D10) 0% GR00T N1.7 0% Pi 0.5 0% Rho 3%
- Manipulation density
- MolmoAct 2 Share of time manipulating, five definitionsJoints moving100.0%End-effector moving98.0%Gripper engaged36.1%Contact force > 5 N83.5%Gripper engaged or in contact89.5% Claude Fable 5.2 (D10) Share of time manipulating, five definitionsJoints moving78.3%End-effector moving56.1%Gripper engaged99.6%Contact force > 5 N46.4%Gripper engaged or in contact99.9% Claude Opus 5.5 (D10) Share of time manipulating, five definitionsJoints moving85.2%End-effector moving68.6%Gripper engaged66.8%Contact force > 5 N92.3%Gripper engaged or in contact97.0% DiT-Flow Share of time manipulating, five definitionsJoints moving99.9%End-effector moving97.8%Gripper engaged35.6%Contact force > 5 N81.2%Gripper engaged or in contact89.1% Gemini 4 Argon (D10) Share of time manipulating, five definitionsJoints moving41.1%End-effector moving30.9%Gripper engaged61.9%Contact force > 5 N66.8%Gripper engaged or in contact91.1% GPT-6 Astra (D10) Share of time manipulating, five definitionsJoints moving63.6%End-effector moving44.7%Gripper engaged27.6%Contact force > 5 N47.5%Gripper engaged or in contact67.0% GR00T N1.7 Share of time manipulating, five definitionsJoints moving100.0%End-effector moving98.9%Gripper engaged41.7%Contact force > 5 N77.4%Gripper engaged or in contact84.7% Pi 0.5 Share of time manipulating, five definitionsJoints moving100.0%End-effector moving99.6%Gripper engaged30.0%Contact force > 5 N78.9%Gripper engaged or in contact86.2% Rho Share of time manipulating, five definitionsJoints moving100.0%End-effector moving98.6%Gripper engaged44.5%Contact force > 5 N78.6%Gripper engaged or in contact86.2%
- SPARC smoothness
- 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. Claude Fable 5.2 (D10) 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. Claude Opus 5.5 (D10) 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. Gemini 4 Argon (D10) 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. GPT-6 Astra (D10) 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. 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. 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.
Episode
Held-out benchmark environment
0:001:02