#sliding-window attention

Sliding-Window Attention: 1 AI articles covering sliding-window attention news, analysis, and research

Articles

Princeton Researcher Proposes Recurrent Looped Transformer (RLT)NEW⭐7

📅 12小时前👁 17 views

Princeton's Recurrent Looped Transformer (RLT) carries decoder state and SWA cache across tokens, with no reset. A design spec for unbounded recurrent depth.