Metraipolis › archive › dc_441c2ccd28dc

Quantization of Large Models with Weight Outliers and Sparsity — arXiv:2611.00286

Source: arXiv (en)
Original: https://arxiv.org/abs/2611.00286
License (as archived): CC0-1.0
Archived: 2026-11-03T11:19:17.625Z
SHA-256 of the archived text: c21eafb6e023627768c10155a00bf1372413595b03330b35f350f8021741f2a1
Stable URL of this page: https://metraipolis.com/documento/dc_441c2ccd28dc

How to cite

Quantization of Large Models with Weight Outliers and Sparsity — arXiv:2611.00286 — arXiv (en), https://arxiv.org/abs/2611.00286 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_441c2ccd28dc (Metraipolis archive dc_441c2ccd28dc, sha256 c21eafb6e023627768c10155a00bf1372413595b03330b35f350f8021741f2a1)

BibTeX

@misc{dc441c2ccd28dc,
  title = {Quantization of Large Models with Weight Outliers and Sparsity — arXiv:2611.00286},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.00286},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_441c2ccd28dc; sha256 c21eafb6e023627768c10155a00bf1372413595b03330b35f350f8021741f2a1},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.00286. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Post-training quantization (PTQ) reduces the memory burden of large model deployment by casting the high precision trained weights to fewer bits. While this promotes memory-efficient inference, it also incurs quantization error that impairs model performance. To compensate for quantization error, the present contribution decomposes the original trained weights into a quantized matrix, a low-rank component, and a sparse matrix that accounts for outlier rows in the weights. Due to analytical intractability and the discrete structure of the decomposition, a simulated annealing search is employed to explore candidate decompositions while avoiding poor local minima. Numerical tests demonstrate the effectiveness of this approach across a broad suite of large language models, including Llama 3.2 1B, Llama 3.2 3B, Llama 3.1 8B, and Qwen2.5 7B, evaluated on natural language generation benchmarks using 4-bit NF4 and FP4 quantizers. The novel method can markedly outperform direct quantization while incurring minimal memory overhead.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2611.00286 — 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 2026-11-03T11:02:37.423Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).