1 paper
Phillip Pramberger, Athanasios Tziouvaras, Shreejith Shanker +1
Sparse attention mechanisms have become increasingly important for transformer models processing long input sequences due to their lower computational and memory complexity compare…