Fetching the paper…

ViG: Linear-complexity Visual Sequence Learning with Gated Linear Attention · Around