Skip to main content
VLA-OPD: Bridging Offline SFT and Online RL for Vision-Language-Action Models via On-Policy Distillation | Signal Canvas | ScienceToStartup