BenchmarkPAW-GEN-10-Agentic
Environmente05Route cable through closed hoops
- Episodes
- Claude Opus 5.5 30 Claude Fable 5.2 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 Opus 5.5 79:52 Claude Fable 5.2 75:55 DiT-Flow 12:25 Gemini 4 Argon 99:25 GPT-6 Astra 82:39 GR00T N1.7 9:21 MolmoAct 2 11:37 Pi 0.5 4:56 Rho 14:21
- Mean duration
- Claude Opus 5.5 159.7 s Claude Fable 5.2 151.8 s DiT-Flow 24.8 s Gemini 4 Argon 198.8 s GPT-6 Astra 165.3 s GR00T N1.7 18.7 s MolmoAct 2 23.2 s Pi 0.5 9.9 s Rho 28.7 s
- Success
- Claude Opus 5.5 0% Claude Fable 5.2 53% DiT-Flow 0% Gemini 4 Argon 17% GPT-6 Astra 0% GR00T N1.7 0% MolmoAct 2 0% Pi 0.5 0% Rho 0%
- Manipulation density
- Claude Opus 5.5 Share of time manipulating, five definitionsJoints moving59.0%End-effector moving43.9%Gripper engaged76.1%Contact force > 5 N78.4%Gripper engaged or in contact93.3% Claude Fable 5.2 Share of time manipulating, five definitionsJoints moving82.4%End-effector moving65.5%Gripper engaged96.8%Contact force > 5 N44.0%Gripper engaged or in contact97.4% DiT-Flow Share of time manipulating, five definitionsJoints moving98.5%End-effector moving93.5%Gripper engaged47.7%Contact force > 5 N86.3%Gripper engaged or in contact91.1% Gemini 4 Argon Share of time manipulating, five definitionsJoints moving56.9%End-effector moving48.0%Gripper engaged23.4%Contact force > 5 N61.5%Gripper engaged or in contact69.1% GPT-6 Astra Share of time manipulating, five definitionsJoints moving67.7%End-effector moving47.4%Gripper engaged50.2%Contact force > 5 N45.1%Gripper engaged or in contact70.8% GR00T N1.7 Share of time manipulating, five definitionsJoints moving98.0%End-effector moving87.2%Gripper engaged65.5%Contact force > 5 N91.8%Gripper engaged or in contact93.4% MolmoAct 2 Share of time manipulating, five definitionsJoints moving96.9%End-effector moving87.9%Gripper engaged66.0%Contact force > 5 N81.6%Gripper engaged or in contact91.8% Pi 0.5 Share of time manipulating, five definitionsJoints moving100.0%End-effector moving98.9%Gripper engaged4.1%Contact force > 5 N85.6%Gripper engaged or in contact85.9% Rho Share of time manipulating, five definitionsJoints moving99.8%End-effector moving97.5%Gripper engaged99.7%Contact force > 5 N83.2%Gripper engaged or in contact99.7%
- SPARC smoothness
- 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. 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. 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…
0:002:36