SOURCE-LINKED INTELLIGENCE

All for 1-Bit: Towards Genuine 1-Bit Post-Training Quantization for LLMs

arXiv · AI, language, vision and robotics · article · Sep 5, 2026 · UTC

Large language models (LLMs) have achieved remarkable progress, yet their massive storage and memory-bandwidth demands still hinder efficient deployment. Weight binarization is a promising solution, but existing binarization-based post-training quantization (PTQ) methods usually far exceed the nominal 1-bit storage target due to hidden overhead. To address this gap, we propose All for 1-Bit (AF1), a genuine 1-bit PTQ framework for LLMs. AF1 comprises two complementary components: (1) Null-space-Aware Binary Factorization (NABF) for improving binary reconstruction through Hessian-aware surrogat

Read original source ↗ Open in workspace

recordType

paper

region

Global