ReGLA: Refining Gated Linear Attention

Open in new window