villa-X: Enhancing Latent Action Modeling in Vision-Language-Action Models

Benchmark Model Rank Results
robot-manipulation-on-simplerenv-widow-xvilla-X (WidowX)#28Average: 0.625
robot-manipulation-on-libero-spatialvilla-X (LIBERO-Spatial)#62Success Rate: 97.5
robot-manipulation-on-libero-objectvilla-X (LIBERO-Object)#82Success Rate: 97.0
robot-manipulation-on-libero-goalvilla-X (LIBERO-Goal)#92Success Rate: 91.5
robot-manipulation-on-libero-longvilla-X (LIBERO-Long)#99Success Rate: 74.5
robot-manipulation-on-liberovilla-X#124Success Rate: 90.1