Fetching the paper…

SARA-RT: Scaling up Robotics Transformers with Self-Adaptive Robust Attention · Around