Metraipolis › archive › dc_3b08fc7df5c7

Faster Than Flash: Exploiting Attention Sparsity for Efficient Long-Context Decoding — arXiv:2609.00097

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.00097
License (as archived)
CC0-1.0
Archived
2026-10-17T06:32:47.262Z
SHA-256 of the archived text
ded06da13d4c92d4a817309982dd64b176d32ba2feda71e2c4934a3ffd038362
Stable URL of this page
https://metraipolis.com/documento/dc_3b08fc7df5c7

How to cite

Faster Than Flash: Exploiting Attention Sparsity for Efficient Long-Context Decoding — arXiv:2609.00097 — arXiv (en), https://arxiv.org/abs/2609.00097 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_3b08fc7df5c7 (Metraipolis archive dc_3b08fc7df5c7, sha256 ded06da13d4c92d4a817309982dd64b176d32ba2feda71e2c4934a3ffd038362)
BibTeX
@misc{dc3b08fc7df5c7,
  title = {Faster Than Flash: Exploiting Attention Sparsity for Efficient Long-Context Decoding — arXiv:2609.00097},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.00097},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_3b08fc7df5c7; sha256 ded06da13d4c92d4a817309982dd64b176d32ba2feda71e2c4934a3ffd038362},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.00097. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

The development of long-context Large Language Models (LLMs) is constrained by the memory bandwidth bottleneck and quadratic complexity of the attention mechanism during decoding. To overcome the inherent trade-offs between the memory overhead of metadata-based metrics and the computational inefficiency of adaptive selection strategies, we present Faster Flash Decoding (FFD), a novel hardware-algorithm co-design framework designed to break the memory wall in long-context decoding. FFD integrates the selector and computer into a fully fused kernel, replacing external metadata indices with content-aware scanning via low-bit quantization. Furthermore, we introduce the top-delta strategy, which dynamically filters blocks to achieve distribution-adaptive sparsity without global synchronization. Offering a training-free and plug-and-play solution, FFD also enables the reuse of scanning results for computation, achieving up to 11.6x kernel-level speedup and scaling to 256K context length, with 2.37x end-to-end throughput improvement. Empirical validation on RULER and LongBench confirms that FFD maintains model accuracy while delivering high-ratio sparsity, with code 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: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.00097 — 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 2026-10-17T06:15:12.742Z 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).