Coverage-Verified Sparse Attention: Closed-Loop Quality Control for Long-Context LLM Decoding

arXiv:2611.10634 · cs.LG, cs.CL · Submitted 2026-11-07 · Read on arXiv

cs.LG, cs.CL

Submitted: 2026-11-07

Updated: 2026-11-07

Code: https://github.com/NVIDIA/kvpress

Terminology

Sources

Related papers