Metraipolis › archive › dc_3102ee10f396
VAR-Q: Tuning-free KV Cache Quantization for Visual Autoregressive Image and Video Generation — arXiv:2611.05380
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.05380
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-05T18:19:49.073Z
- SHA-256 of the archived text
c7c48797491681421501b2f304a1fa1d2e9517892106e52db49c9f440863ccbe- Stable URL of this page
https://metraipolis.com/documento/dc_3102ee10f396
How to cite
VAR-Q: Tuning-free KV Cache Quantization for Visual Autoregressive Image and Video Generation — arXiv:2611.05380 — arXiv (en), https://arxiv.org/abs/2611.05380 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_3102ee10f396 (Metraipolis archive dc_3102ee10f396, sha256 c7c48797491681421501b2f304a1fa1d2e9517892106e52db49c9f440863ccbe)
BibTeX
@misc{dc3102ee10f396,
title = {VAR-Q: Tuning-free KV Cache Quantization for Visual Autoregressive Image and Video Generation — arXiv:2611.05380},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.05380},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_3102ee10f396; sha256 c7c48797491681421501b2f304a1fa1d2e9517892106e52db49c9f440863ccbe},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.05380. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Visual autoregressive generation spans two dominant paradigms: next-scale prediction for images and video, and next-frame prediction for video. Both decode block-by-block rather than token-by-token, yielding KV caches whose statistics diverge from those of language models. At scale, KV memory exceeds the weights and limits large-scale deployment caused by a large amount of visual tokens. Our analysis of these caches reveals three findings. First, a small set of massive-value channels carries the functional load: ablating these outlier channels collapses generation, whereas ablating a matched set of inlier channels leaves quality intact. Second, the outliers organize along two axes, feature and block, with magnitudes persistent within each cell and shifting across cells. Third, this structure arises from two independent mechanisms: projection bias and RoPE-induced scattering. Guided by these findings, we propose VAR-Q, a tuning-free KV-cache quantization framework that groups along the same two axes. The generation loop provides block boundaries for free, removing the need for outlier detection or calibration. Paired with bit-packing kernels for runtime memory reduction, VAR-Q delivers low-bit quantization across nine models spanning both paradigms, cutting KV memory by up to 78.5% with negligible quality degradation and enabling 6x larger inference batches. Tema degli esploratori: calcolo e infrastrutture (elenco arXiv cs.PF, nuovi lavori del giorno). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Performance (cs.PF). Pagina del lavoro: https://arxiv.org/abs/2611.05380 — 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 2026-11-05T18:13:48.991Z dalla pagina https://arxiv.org/list/cs.PF/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).