Simulation

Synthetic-Data RL for VLA Models

Training vision-language-action models for complex manipulation without human demonstrations.