Simulation
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Simulation Jaco robot
36Success Rate
4
Simulation Jaco robot
38Success Rate
4
Simulation Panda robot
40Success Rate
4
Simulation 2
98Mean F1 Score
4
Simulation 1
96.9Mean F1 Score
4
Simulation cluttered forest, 60m × 60m × 5m
13.63Avg Velocity (m/s)
4
Simulation 15m × 120m × 4m
22.1Avg Velocity (m/s)
4
Simulation Non-prior Crossing
33.1TPR
4
Simulation Non-prior One-direction
54.7TPR
4
Simulation Prior One-direction
41.2TPR
4
Simulation
93.3Success Rate
4
Simulation
22Cleaned Stains per Patch
4
Simulation Held-out Tasks
45Success Rate
3
Simulation Seen Tasks
61Success Rate
3
Simulation Scenario 1
99.5Specificity
3
Simulation 1
0.016MAE (Translation) (m)
3
Simulation Reasoning
80.89Success Rate
3
Simulation Long-Horizon
67.08Success Rate
3
Simulation Short-Horizon
78.22Success Rate
3
Simulation Case 1
0.101Miss Distance (m)
3
Simulation Case 5 Strong overlap n=300
75.3True Positive Rate (TPR)
3
Simulation Case 2 Weak main n=300
100TPR
3
Simulation Hidden confounding size 1000 (test)
2.369Mean CATE Bias
3
Simulation N=2,000 tasks seed 42 (test)
179Mean Latency (ms)
3
Simulation
40.22Success Rate
3