BenchmarkPAW-GEN-10
Environmente02Datum-corner alignment
- Episodes
- DiT-Flow 30 Claude Fable 5.2 (D10) 30 Claude Opus 5.5 (D10) 30 Gemini 4 Argon (D10) 30 GPT-6 Astra (D10) 30 GR00T N1.7 30 MolmoAct 2 30 Pi 0.5 30 Rho 30
- Total duration
- DiT-Flow 7:47 Claude Fable 5.2 (D10) 20:34 Claude Opus 5.5 (D10) 69:54 Gemini 4 Argon (D10) 26:54 GPT-6 Astra (D10) 48:26 GR00T N1.7 8:02 MolmoAct 2 6:41 Pi 0.5 7:03 Rho 10:24
- Mean duration
- DiT-Flow 15.6 s Claude Fable 5.2 (D10) 41.1 s Claude Opus 5.5 (D10) 139.8 s Gemini 4 Argon (D10) 53.8 s GPT-6 Astra (D10) 96.9 s GR00T N1.7 16.1 s MolmoAct 2 13.4 s Pi 0.5 14.1 s Rho 20.8 s
- Success
- DiT-Flow 20% Claude Fable 5.2 (D10) 83% Claude Opus 5.5 (D10) 47% Gemini 4 Argon (D10) 27% GPT-6 Astra (D10) 93% GR00T N1.7 0% MolmoAct 2 60% Pi 0.5 57% Rho 60%
- Manipulation density
- DiT-Flow Share of time manipulating, five definitionsJoints moving99.3%End-effector moving86.7%Gripper engaged0.0%Contact force > 5 N72.9%Gripper engaged or in contact72.9% Claude Fable 5.2 (D10) Share of time manipulating, five definitionsJoints moving90.4%End-effector moving80.9%Gripper engaged43.1%Contact force > 5 N86.6%Gripper engaged or in contact89.4% Claude Opus 5.5 (D10) Share of time manipulating, five definitionsJoints moving84.6%End-effector moving69.6%Gripper engaged0.0%Contact force > 5 N82.1%Gripper engaged or in contact82.1% Gemini 4 Argon (D10) Share of time manipulating, five definitionsJoints moving52.8%End-effector moving47.0%Gripper engaged0.0%Contact force > 5 N75.6%Gripper engaged or in contact75.6% GPT-6 Astra (D10) Share of time manipulating, five definitionsJoints moving84.5%End-effector moving70.1%Gripper engaged0.0%Contact force > 5 N82.2%Gripper engaged or in contact82.2% GR00T N1.7 Share of time manipulating, five definitionsJoints moving99.8%End-effector moving96.5%Gripper engaged0.0%Contact force > 5 N77.5%Gripper engaged or in contact77.5% MolmoAct 2 Share of time manipulating, five definitionsJoints moving100.0%End-effector moving96.1%Gripper engaged0.0%Contact force > 5 N72.2%Gripper engaged or in contact72.2% Pi 0.5 Share of time manipulating, five definitionsJoints moving99.7%End-effector moving96.4%Gripper engaged0.0%Contact force > 5 N70.7%Gripper engaged or in contact70.7% Rho Share of time manipulating, five definitionsJoints moving99.9%End-effector moving95.4%Gripper engaged0.0%Contact force > 5 N74.0%Gripper engaged or in contact74.0%
- SPARC smoothness
- 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. 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. 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. 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
Held-out benchmark environment
0:000:17