Sparser is Faster and Less is More: Efficient Sparse Attention for Long-Range Transformers Paper • 2406.16747 • Published Jun 24, 2024 • 18