A Unified Hessian-Guided Framework for Post-Training Quantization
arXiv:2610.40345v1 Announce Type: cross Abstract: Low-bit quantization enables efficient deployment of large language models by reducing memory footprint and computational cost. Achieving high accuracy at low precision requires selecting quantization ranges, linear transformations, and weight rounding decisions. Existing approaches typically optimize these components independently using local reconstruction objectives, which fail to account for the varying impact of quantization errors on downs…