Accelerating Long-Context Decoding in Large-Scale Hybrid LLMs Without Retraining

arXiv:2611.06133 · cs.CL, cs.PF · Submitted 2026-11-04 · Read on arXiv

cs.CL, cs.PF

Submitted: 2026-11-04

Updated: 2026-11-04

Code: https://github.com/jespb/CombinedSparseAttention

Terminology

Sources

Related papers