Based: Simple linear attention language models by from Hacker News on 2024-03-05 00:15 (#6K3BA) Comments