← back
arXivAles Kubicek, Andrea Caciolai, Liangzu Peng, Sukrit Sumant, Benjamin Hoover, Fraser Greenlee, Ruairidh M. Battleday, Radim Urban, Torsten Hoefler, James C. R. WhittingtonFri, Oct 30, 2026, 10:59 AM PDT
score 15.3

Sparse Attention Method Makes Million-Token Contexts Affordable

Original: Latent Sparse Attention: Towards Subquadratic Ultra-Long Context Attention

Source: arxiv.org ↗

Writing ELI5 summary…