Simple linear attention language models balance the recall-throughput tradeoff Paper • 2402.18668 • Published Feb 28, 2024 • 19
