Metraipolis › archive › dc_f45201bb9580
A Unified Hessian-Guided Framework for Post-Training Quantization — arXiv:2610.40345
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.40345
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-31T06:14:52.563Z
- SHA-256 of the archived text
3197b544aa978b3278d7aff2289a5622f15a9063c966bc8f5309195a9e4d2d0e- Stable URL of this page
https://metraipolis.com/documento/dc_f45201bb9580
How to cite
A Unified Hessian-Guided Framework for Post-Training Quantization — arXiv:2610.40345 — arXiv (en), https://arxiv.org/abs/2610.40345 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_f45201bb9580 (Metraipolis archive dc_f45201bb9580, sha256 3197b544aa978b3278d7aff2289a5622f15a9063c966bc8f5309195a9e4d2d0e)
BibTeX
@misc{dcf45201bb9580,
title = {A Unified Hessian-Guided Framework for Post-Training Quantization — arXiv:2610.40345},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.40345},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_f45201bb9580; sha256 3197b544aa978b3278d7aff2289a5622f15a9063c966bc8f5309195a9e4d2d0e},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.40345. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
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 downstream predictions. Direct optimization of task-level metrics would provide a more faithful objective, but is challenging due to the discontinuous nature of quantization operations and the high computational cost of evaluating their downstream effects. In this work, we derive a Hessian-guided approximation of the output change induced by joint weight and activation quantization. Based on a tractable Fisher approximation, we formulate a unified post-training quantization framework that enables the joint optimization of quantization ranges, linear transformations, and weight rounding under a common output-sensitivity objective. The resulting method remains practical for modern large language models and consistently outperforms state-of-the-art transformation and rounding methods across both channel-wise and block-wise quantization settings. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2610.40345 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-31T06:07:08.768Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).