AI RESEARCH
SAMoE-VLA: A Scene Adaptive Mixture-of-Experts Vision-Language-Action Model for Autonomous Driving
arXiv CS.CV
•
ArXi:2603.08113v1 Announce Type: new Recent advances in Vision-Language-Action (VLA) models have shown promising capabilities in autonomous driving by leveraging the understanding and reasoning strengths of Large Language Models(LLMs). However, our empirical analysis reveals that directly applying existing token-level MoE mechanisms--which are inherited from LLM architectures--to VLA models results in unstable performance and safety degradation in autonomous driving, highlighting a misalignment between token-based expert specialization and scene-level decision-making.