F1: A Vision-Language-Action Model Bridging Understanding and Generation to Actions

Benchmark Model Rank Results
robot-manipulation-on-libero-spatialF1 (pretrained) (LIBERO-Spatial)#48Success Rate: 98.2
robot-manipulation-on-libero-objectF1 (pretrained) (LIBERO-Object)#76Success Rate: 97.8
robot-manipulation-on-libero-goalF1 (pretrained) (LIBERO-Goal)#78Success Rate: 95.4
robot-manipulation-on-libero-longF1 (pretrained) (LIBERO-Long)#79Success Rate: 91.3
robot-manipulation-on-liberoF1 (pretrained)#102Success Rate: 95.7