PerturBot: Breaking Shortcut Priors in Vision-Language-Action Models with Perturbative Training
Open paper
Benchmark
Model
Rank
Results
robot-manipulation-on-robotwin-2-0-hard-50-tasks
Perturbot (extra data)
#29
Success Rate: 81.9
robot-manipulation-on-robotwin-2-0-easy-50-tasks
Perturbot (extra data)
#31
Success Rate: 85.6
Rank counts only results with a code link.