Preprint
Aug 2026
CubicQuant: Parametric Non-Uniform Codebooks for High-Throughput LLM Inference with 1-8-Bit Weights
This work introduces CubicQuant, a parametric non-uniform scalar format that preserves a dense integer code stream while adapting reconstruction levels within each weight group, and establishes its representational promise and direct executability.
Xu Gao
· 0 citations