BenchmarkPAW-GEN-10-Agentic
Environmente01Sort screws into bins
- Episodes
- Claude Fable 5.2 30 Claude Opus 5.5 30 DiT-Flow 30 Gemini 4 Argon 30 GPT-6 Astra 30 GR00T N1.7 30 MolmoAct 2 30 Pi 0.5 30 Rho 30
- Total duration
- Claude Fable 5.2 49:05 Claude Opus 5.5 55:27 DiT-Flow 13:02 Gemini 4 Argon 90:32 GPT-6 Astra 56:47 GR00T N1.7 11:48 MolmoAct 2 16:51 Pi 0.5 14:26 Rho 12:00
- Mean duration
- Claude Fable 5.2 98.2 s Claude Opus 5.5 110.9 s DiT-Flow 26.1 s Gemini 4 Argon 181.1 s GPT-6 Astra 113.6 s GR00T N1.7 23.6 s MolmoAct 2 33.7 s Pi 0.5 28.9 s Rho 24.0 s
- Success
- Claude Fable 5.2 87% Claude Opus 5.5 73% DiT-Flow 0% Gemini 4 Argon 27% GPT-6 Astra 40% GR00T N1.7 0% MolmoAct 2 0% Pi 0.5 0% Rho 0%
- Manipulation density
- Claude Fable 5.2 Share of time manipulating, five definitionsJoints moving81.2%End-effector moving57.3%Gripper engaged25.9%Contact force > 5 N50.7%Gripper engaged or in contact58.8% Claude Opus 5.5 Share of time manipulating, five definitionsJoints moving90.0%End-effector moving71.1%Gripper engaged53.5%Contact force > 5 N62.7%Gripper engaged or in contact80.2% DiT-Flow Share of time manipulating, five definitionsJoints moving95.2%End-effector moving84.5%Gripper engaged63.4%Contact force > 5 N62.8%Gripper engaged or in contact87.3% Gemini 4 Argon Share of time manipulating, five definitionsJoints moving57.7%End-effector moving39.2%Gripper engaged92.6%Contact force > 5 N84.9%Gripper engaged or in contact98.6% GPT-6 Astra Share of time manipulating, five definitionsJoints moving78.2%End-effector moving51.2%Gripper engaged33.8%Contact force > 5 N70.6%Gripper engaged or in contact78.5% GR00T N1.7 Share of time manipulating, five definitionsJoints moving100.0%End-effector moving99.1%Gripper engaged61.2%Contact force > 5 N66.7%Gripper engaged or in contact91.0% MolmoAct 2 Share of time manipulating, five definitionsJoints moving99.9%End-effector moving95.4%Gripper engaged52.1%Contact force > 5 N70.7%Gripper engaged or in contact81.5% Pi 0.5 Share of time manipulating, five definitionsJoints moving99.2%End-effector moving96.5%Gripper engaged30.8%Contact force > 5 N60.8%Gripper engaged or in contact70.5% Rho Share of time manipulating, five definitionsJoints moving97.0%End-effector moving84.6%Gripper engaged26.6%Contact force > 5 N67.4%Gripper engaged or in contact71.4%
- SPARC smoothness
- Claude Fable 5.2 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 Opus 5.5 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. 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 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 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. 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
Loading the replay…
0:001:21