EfficientVLA: Training-Free Acceleration and Compression for Vision-Language-Action Models Paper • 2506.10100 • Published Jun 11, 2025 • 9
