Skip to content

JanusQuant: Accurate and Efficient 2-bit KV Cache Quantization for Long-Context Inference.

Chengyu Sun, Yaqi Xia, Hulin Wang, Donglin Yang, Xiaobo Zhou, Dazhao Cheng

VenueBPPoPP
Year2026
ProceedingsPPoPP

Browse the full PPoPP paper archive.