Metraipolis › archive › dc_fac98e193e73
OCGQuant: Outlier-Companion Grouping for NVFP4 Quantization — arXiv:2609.00066
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.00066
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T06:32:41.589Z
- SHA-256 of the archived text
49a435ff70c45e400893ca1f9fe7efa0f80502f310fb84f7da80762e49a7aeac- Stable URL of this page
https://metraipolis.com/documento/dc_fac98e193e73
How to cite
OCGQuant: Outlier-Companion Grouping for NVFP4 Quantization — arXiv:2609.00066 — arXiv (en), https://arxiv.org/abs/2609.00066 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_fac98e193e73 (Metraipolis archive dc_fac98e193e73, sha256 49a435ff70c45e400893ca1f9fe7efa0f80502f310fb84f7da80762e49a7aeac)
BibTeX
@misc{dcfac98e193e73,
title = {OCGQuant: Outlier-Companion Grouping for NVFP4 Quantization — arXiv:2609.00066},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.00066},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_fac98e193e73; sha256 49a435ff70c45e400893ca1f9fe7efa0f80502f310fb84f7da80762e49a7aeac},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.00066. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
NVFP4 is an efficient microscaling format for low-bit inference, but activation outliers can still degrade quantization accuracy within NVFP4 blocks. Within each quantization block, large activations can dominate the block scale, increasing the quantization error of the remaining values sharing the same scale. Existing post-training quantization (PTQ) methods mitigate outlier errors through strategies such as mixed precision, rotation, or residual compensation, but these approaches are either not specifically tailored to NVFP4 or introduce additional computation. In this work, we revisit NVFP4 from a channel-grouping perspective and define the reducible error incurred by remaining block values under the scale set by the block maximum as Collateral Quantization Error. Based on this insight, we propose OCGQuant, a post-training quantization method centered on Outlier-Companion Grouping (OCG), which adaptively pairs outlier channels with low-magnitude companion channels to improve NVFP4 activation block composition. Experiments on Llama3 and Qwen3 show that OCGQuant achieves the lowest WikiText-2 perplexity and highest average downstream accuracy among evaluated PTQ methods, while maintaining prefill speedup close to RTN and matching its peak decoding memory. Code is available at this https URL . Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-02 in cs.AI). Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI); Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2609.00066 — 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 2026-10-17T06:15:12.730Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).