CogVLA: Cognition-Aligned Vision-Language-Action Model via Instruction-Driven Routing & Sparsification 事件

CogVLA: Cognition-Aligned Vision-Language-Action Model via Instruction-Driven Routing & Sparsification arXiv:2508.21046v3 Announce Type: replace Abstract: Recent Vision-Language-Action (VLA) models built on pre-trained Vision-Language Models (VLMs) require extensive post-training, resulting in high computational overhead that limits scalability and deployment.We propose CogVLA, a Cognition-Aligned Vision-Language-Action framework that leverages instruction-driven routing and sparsification to i