Multi-Future Drafting：混合块草拟器实现树验证推测解码加速
Multi-Future Drafting: Mixture Block Drafters for Tree-Verified Speculative Decoding
AI 导读
arXiv 论文提出 Multi-Future Drafting（MFD）方法，用 K 个可学习的模式嵌入在一次前向中草拟 K 个连贯块，并配合混合 DDTree 让验证器无损选出目标模型采纳的未来分支，仅需 K+1 个额外向量。
正文 · 原文
Abstract:Block drafters such as DFlash draft a block of future tokens in one forward pass, and draft trees such as DDTree verify the prefixes ranked highest by the product of the drafter's per-position distributions. Under tree verification, the expected number of accepted tokens is the probability mass the tree covers. We show that a product misallocates it: when the continuation can follow several futures, it splices them into chimeras, prefixes stitched together from different futures that the target never produces, and gives them as much budget as the futures themselves. Barring ties, no product-ranked tree can hold two futures that disagree in $m$ positions without $2^{m-1}-1$ chimeras, however well the drafter is trained, and in practice chimeras make up 27-33% of a 64-node tree. We propose Multi-Future Drafting (MFD), which restores the coupling with a latent variable instead of the chain rule: $K$ learned mode embeddings at the drafter's input draft $K$ coherent blocks in the same pass, and a mixture DDTree lets the verifier pick the future the target takes, losslessly and with $K+1$ extra vectors. On four Qwen targets from 4B to 27B parameters, at batch size 1, MFD reaches up to 7.8$\times$ speedup, 5-8% faster than the best single-mode tree trained identically and, on Qwen3-4B and Qwen3-8B, 9-16% faster than the released DARTree trained on the same prompts.
| Subjects: | Artificial Intelligence (cs.AI); Machine Learning (cs.LG) |
| Cite as: | arXiv:2611.13017 [cs.AI] |
| (or arXiv:2611.13017v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2611.13017 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Haoyu Huang [view email]
[v1]
Mon, 9 Nov 2026 05:15:08 UTC (185 KB)
来源：arXiv cs.AI · arxiv.org