Metraipolis › archive › dc_39740ad95bf7
SpecFold: Folding Multi-Branch Redundancy for Faster Speculative Decoding in Diffusion Language Models — arXiv:2610.04875
- Source
- arXiv
- Original
- https://arxiv.org/abs/2610.04875
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-06T05:24:51.760Z
- SHA-256 of the archived text
b87e6b26a91a222094435ca025a5d82ceaf9b8c7d9d34145815a8d9abeee2a2a- Stable URL of this page
https://metraipolis.com/documento/dc_39740ad95bf7
How to cite
SpecFold: Folding Multi-Branch Redundancy for Faster Speculative Decoding in Diffusion Language Models — arXiv:2610.04875 — arXiv, https://arxiv.org/abs/2610.04875 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_39740ad95bf7 (Metraipolis archive dc_39740ad95bf7, sha256 b87e6b26a91a222094435ca025a5d82ceaf9b8c7d9d34145815a8d9abeee2a2a)
BibTeX
@misc{dc39740ad95bf7,
title = {SpecFold: Folding Multi-Branch Redundancy for Faster Speculative Decoding in Diffusion Language Models — arXiv:2610.04875},
howpublished = {arXiv},
url = {https://arxiv.org/abs/2610.04875},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_39740ad95bf7; sha256 b87e6b26a91a222094435ca025a5d82ceaf9b8c7d9d34145815a8d9abeee2a2a},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.04875. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Diffusion large language models (DLLMs) generate text through iterative block denoising, and multi-branch speculative decoding accelerates this process by verifying a main branch together with multiple draft branches in a single forward pass. While prior DLLM acceleration methods primarily exploit temporal redundancy across denoising steps, we identify a complementary redundancy axis within each speculative verification step: multi-branch computational redundancy. During speculative verification, draft branches inherit most tokens from their parents while unmasking a small set of additional positions, causing large portions of hidden states to remain highly similar across branches. We propose SpecFold, an algorithm-system co-design that exploits this multi-branch redundancy to reduce the cost of multi-branch speculative verification. Algorithmically, SpecFold performs token-level residual gating and selectively reuses parent computation through folded attention and FFN while preserving residual hidden states. Systemically, a Triton kernel implementation translates this fine-grained reuse into end-to-end throughput gains through efficient sparse multi-branch execution. SpecFold is orthogonal to temporal caching and compatible with existing DLLM speculation strategies. Across two DLLM families, five models, and five standard benchmarks, SpecFold achieves up to 1.64x throughput over Spiffy and up to 1.99x over vanilla decoding, while maintaining comparable task performance. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Artificial Intelligence (cs.AI) ; Computation and Language (cs.CL); Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2610.04875 — 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 2026-10-06T05:24:05.528Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).