STARE-VLA

Progressive Stage-Aware Reinforcement for Fine-Tuning Vision-Language-Action Models