Metraipolis › archive › dc_27c90649cfaf

SparseGDN: Chunk Sparse in Linear Attention — arXiv:2610.41420

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.41420
License (as archived)
CC0-1.0
Archived
2026-11-02T07:10:40.474Z
SHA-256 of the archived text
9dbdf398c9b8267f3dc04df8e5a5b99667b0df02ed59a41aca13839e954b810d
Stable URL of this page
https://metraipolis.com/documento/dc_27c90649cfaf

How to cite

SparseGDN: Chunk Sparse in Linear Attention — arXiv:2610.41420 — arXiv (en), https://arxiv.org/abs/2610.41420 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_27c90649cfaf (Metraipolis archive dc_27c90649cfaf, sha256 9dbdf398c9b8267f3dc04df8e5a5b99667b0df02ed59a41aca13839e954b810d)
BibTeX
@misc{dc27c90649cfaf,
  title = {SparseGDN: Chunk Sparse in Linear Attention — arXiv:2610.41420},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.41420},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_27c90649cfaf; sha256 9dbdf398c9b8267f3dc04df8e5a5b99667b0df02ed59a41aca13839e954b810d},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.41420. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Gated Delta Networks (GDNs) offer linear-time sequence modeling, yet long-context prefill remains costly due to dense intra-chunk computation, recurrent-state updates, and intermediate state storage. We introduce SparseGDN, a hardware-aware framework that combines structured intra-chunk approximation with input-dependent recurrent-event selection while producing outputs for every token. A block-adjacent solver reduces local Delta-solve computation, while a dynamic selector allocates recurrent-update budgets across head pairs. The GPU implementation materializes updates only for selected events and reconstructs token outputs using compact state snapshots and group-local interactions, reducing computation and memory traffic. We evaluate SparseGDN on Qwen3.5-4B and Qwen3.5-9B across RULER tasks and context lengths from 4K to 128K. Retaining approximately 29% of recurrent events on average, SparseGDN achieves 1.50-1.94x GDN-core prefill speedups over native FLA kernels, including online selection overhead.

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); Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2610.41420 — 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 2026-11-02T07:06:59.441Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).