#artificial intelligence
Feb 2026
SCALE: Self-uncertainty Conditioned Adaptive Looking and Execution for Vision-Language-Action Models
Experiments on simulated and real-world benchmarks demonstrate that SCALE improves state-of-the-art VLAs and outperforms existing TTS methods while maintaining single-pass efficiency.
Hyeonbeom Choi, Daechul Ahn, Youhan Lee et al.
· arXiv.org · 3 citations