Kuaishou Tech
May 14, 2026 · Artificial Intelligence
Open‑Source Kwai Summary Attention (KSA): A Sequence‑Compression Mechanism for Long‑Context Inference
KSA inserts learnable summary tokens to compress KV cache by a factor of eight, enabling accurate long‑context retrieval with far lower memory and compute costs, and it consistently outperforms full‑attention and other hybrid methods on large‑scale benchmarks.
KSAKV cache reductionLarge Language Models
0 likes · 13 min read