Metraipolis › archive › dc_5cc6650dcb2d
BitNest: Bit-Nested Speculative Decoding for Memory-Efficient LLM Inference Acceleration — arXiv:2610.02800
- Source
- arXiv
- Original
- https://arxiv.org/abs/2610.02800
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-05T22:49:14.243Z
- SHA-256 of the archived text
6a4db01f3aa1d7cce7e0634f3d04321c5118336f11bb76f09d30b84c20188146- Stable URL of this page
https://metraipolis.com/documento/dc_5cc6650dcb2d
How to cite
BitNest: Bit-Nested Speculative Decoding for Memory-Efficient LLM Inference Acceleration — arXiv:2610.02800 — arXiv, https://arxiv.org/abs/2610.02800 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_5cc6650dcb2d (Metraipolis archive dc_5cc6650dcb2d, sha256 6a4db01f3aa1d7cce7e0634f3d04321c5118336f11bb76f09d30b84c20188146)
BibTeX
@misc{dc5cc6650dcb2d,
title = {BitNest: Bit-Nested Speculative Decoding for Memory-Efficient LLM Inference Acceleration — arXiv:2610.02800},
howpublished = {arXiv},
url = {https://arxiv.org/abs/2610.02800},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_5cc6650dcb2d; sha256 6a4db01f3aa1d7cce7e0634f3d04321c5118336f11bb76f09d30b84c20188146},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.02800. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Speculative decoding accelerates autoregressive generation by using a lightweight draft to propose multiple tokens for parallel verification. However, existing methods often require an additional draft model or weight representation, introducing non-negligible memory overhead on resource-constrained devices. Self-speculative approaches reduce this overhead, yet still face trade-offs between draft quality, target quality, and storage efficiency. We propose BitNest, a bit-nested speculative decoding framework that embeds a low-precision draft directly into the higher-precision target representation. Instead of deriving a draft from a predefined target, BitNest first constructs a strong low-precision base and then recovers the higher-precision target through residual refinement, enabling both models to share a single physical weight representation. BitNest further extends this progressive-precision design to the KV cache for long-context inference. Across multiple 7B--8B edge-friendly LLMs and diverse workloads, BitNest achieves an average speculative acceptance rate of 95.2% while closely preserving higher-precision model quality, and delivers 1.48--1.61x end-to-end speedup over FP16 autoregressive decoding. On the LLaMA models supported by all representative self-speculative baselines, BitNest also achieves consistently competitive or higher decoding speedup. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2610.02800 — 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 2026-10-05T22:45:55.631Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).