DeeR-VLA: Dynamic Inference of Multimodal Large Language Models for Efficient Robot Execution

Open in new window