BenchmarkPAW-GEN-10-Agentic
Environmente03Pour screws into the funnel
- Episodes
- GPT-6 Astra 30 Claude Fable 5.2 30 Claude Opus 5.5 30 DiT-Flow 30 Gemini 4 Argon 30 GR00T N1.7 30 MolmoAct 2 30 Pi 0.5 30 Rho 30
- Total duration
- GPT-6 Astra 35:59 Claude Fable 5.2 26:44 Claude Opus 5.5 24:40 DiT-Flow 3:22 Gemini 4 Argon 19:57 GR00T N1.7 3:25 MolmoAct 2 4:53 Pi 0.5 8:51 Rho 1:29
- Mean duration
- GPT-6 Astra 72.0 s Claude Fable 5.2 53.5 s Claude Opus 5.5 49.3 s DiT-Flow 6.7 s Gemini 4 Argon 39.9 s GR00T N1.7 6.8 s MolmoAct 2 9.8 s Pi 0.5 17.7 s Rho 3.0 s
- Success
- GPT-6 Astra 60% Claude Fable 5.2 93% Claude Opus 5.5 77% DiT-Flow 0% Gemini 4 Argon 93% GR00T N1.7 0% MolmoAct 2 3% Pi 0.5 7% Rho 0%
- Manipulation density
- GPT-6 Astra Share of time manipulating, five definitionsJoints moving73.4%End-effector moving53.0%Gripper engaged56.5%Contact force > 5 N71.2%Gripper engaged or in contact77.6% Claude Fable 5.2 Share of time manipulating, five definitionsJoints moving90.7%End-effector moving69.9%Gripper engaged58.1%Contact force > 5 N43.8%Gripper engaged or in contact74.8% Claude Opus 5.5 Share of time manipulating, five definitionsJoints moving90.6%End-effector moving79.9%Gripper engaged62.5%Contact force > 5 N61.8%Gripper engaged or in contact79.3% DiT-Flow Share of time manipulating, five definitionsJoints moving99.7%End-effector moving96.7%Gripper engaged49.3%Contact force > 5 N59.2%Gripper engaged or in contact80.6% Gemini 4 Argon Share of time manipulating, five definitionsJoints moving88.2%End-effector moving80.6%Gripper engaged64.3%Contact force > 5 N80.2%Gripper engaged or in contact89.0% GR00T N1.7 Share of time manipulating, five definitionsJoints moving100.0%End-effector moving99.5%Gripper engaged67.9%Contact force > 5 N61.6%Gripper engaged or in contact84.0% MolmoAct 2 Share of time manipulating, five definitionsJoints moving100.0%End-effector moving99.5%Gripper engaged38.1%Contact force > 5 N45.4%Gripper engaged or in contact62.0% Pi 0.5 Share of time manipulating, five definitionsJoints moving100.0%End-effector moving99.7%Gripper engaged84.9%Contact force > 5 N40.7%Gripper engaged or in contact91.5% Rho Share of time manipulating, five definitionsJoints moving100.0%End-effector moving98.2%Gripper engaged42.8%Contact force > 5 N54.5%Gripper engaged or in contact64.4%
- SPARC smoothness
- 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. 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. 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…
0:001:17