Metraipolis › archive › dc_059fc6c0df60
C3: Speech-Guided KV Retrieval for Long-Context SpeechLLM Inference — arXiv:2611.00333
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.00333
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-03T10:34:07.771Z
- SHA-256 of the archived text
3f5c364d7474d70380c47d2b52ebc785dd48e55df554ee650f32257a33dcdf0f- Stable URL of this page
https://metraipolis.com/documento/dc_059fc6c0df60
How to cite
C3: Speech-Guided KV Retrieval for Long-Context SpeechLLM Inference — arXiv:2611.00333 — arXiv (en), https://arxiv.org/abs/2611.00333 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_059fc6c0df60 (Metraipolis archive dc_059fc6c0df60, sha256 3f5c364d7474d70380c47d2b52ebc785dd48e55df554ee650f32257a33dcdf0f)
BibTeX
@misc{dc059fc6c0df60,
title = {C3: Speech-Guided KV Retrieval for Long-Context SpeechLLM Inference — arXiv:2611.00333},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.00333},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_059fc6c0df60; sha256 3f5c364d7474d70380c47d2b52ebc785dd48e55df554ee650f32257a33dcdf0f},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.00333. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Speech language models encode audio at a much higher rate than text, causing the KV cache to dominate the memory, bandwidth, and decoding cost of long-form generation. KV retrieval can reduce per-step access by attending to only a query-dependent subset of the cache, but existing methods are designed primarily for text and do not account for the structure of audio attention. We first show that retrieval errors may compound autoregressively into generation drift. We then analyze full audio attention during long-form generation and identify three properties relevant to retrieval: unequal attention concentration across query heads, non-contiguous audio support, and temporal continuity across decoding steps. Guided by these observations, we propose $C^3$(Cluster-Compete-Compete), which organizes audio KVs in a hierarchical key-space index, lets query heads sharing a KV head compete for a common retrieval budget, and adaptively reallocates retrieval resolution through paired split-merge updates. Across long-form speech recognition, translation, and summarization on four SpeechLLMs, $C^3$ achieves the strongest overall performance among retrieval methods under comparable final-KV access budgets, with particularly large gains in the tight-budget regime, while improving long-context decoding efficiency and accelerator-memory scaling. 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); Sound (cs.SD); Audio and Speech Processing (eess.AS). Pagina del lavoro: https://arxiv.org/abs/2611.00333 — 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 2026-11-03T10:29:48.683Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).