Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Simulation Benchmark

Benchmarks

Task NameDataset NameSOTA ResultTrend
Long-horizon manipulationSimulation Benchmark
Retrieve Object Success Rate (Spatial)68
8
Goal-conditioned NavigationSimulation Benchmark
Goal Reached Rate61
8
Closed-loop long-horizon navigationsimulation benchmark
NIR21.2
7
Functional Tool-UseSimulation Benchmark unseen tools
SR (Scoop S01)30
4
Overall manipulation successSimulation Benchmark Overall OOD
Success Rate71
4
Overall manipulation successSimulation Benchmark Overall ID
Success Rate63
4
PrepareTable manipulationSimulation Benchmark PrepareTable OOD
Success Rate72
4
PrepareTable manipulationSimulation Benchmark PrepareTable S1
Success Rate60
4
HangMugTree manipulationSimulation Benchmark HangMugTree OOD
Success Rate60
4
PourInBowl manipulationSimulation Benchmark PourInBowl OOD
Success Rate80
4
PourInBowl manipulationSimulation Benchmark PourInBowl S2
Success Rate56
4
Active Exploration and GraspingSimulation Benchmark Level 4
TFT (s)131.4
4
Active Exploration and GraspingSimulation Benchmark Level 3
Time to First Touch (s)95
4
Active Exploration and GraspingSimulation Benchmark Level 2
TFT (s)133.9
4
Active Exploration and GraspingSimulation Benchmark Level 1
TFT (s)83.5
4
Showing 15 of 15 rows