SpecTr-GBV: Multi-Draft Block Verification Accelerating Speculative Decoding

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Lin, Yijun, Sheng, Jinhao, Cai, Qingyue, Zhou, Feng
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913070668840960
author Lin, Yijun
Sheng, Jinhao
Cai, Qingyue
Zhou, Feng
author_facet Lin, Yijun
Sheng, Jinhao
Cai, Qingyue
Zhou, Feng
contents Autoregressive language models suffer from high inference latency due to their sequential decoding nature. Speculative decoding (SD) mitigates this by employing a lightweight draft model to propose candidate tokens, which are selectively verified by a larger target model. While existing methods either adopt multi-draft strategies to increase acceptance rates or block verification techniques to jointly verify multiple tokens, they remain limited by treating these improvements in isolation. In this work, we propose SpecTr-GBV, a novel SD method that unifies multi-draft and greedy block verification (GBV) into a single framework. By formulating the verification step as an optimal transport problem over draft and target token blocks, SpecTr-GBV improves both theoretical efficiency and empirical performance. We theoretically prove that SpecTr-GBV achieves the optimal expected acceptance length physically attainable within the framework of i.i.d. draft generation, and this bound improves as the number of drafts increases. Empirically, we evaluate SpecTr-GBV across five datasets and four baselines. Our method achieves superior speedup and significantly higher block efficiency while preserving output quality. In addition, we perform comprehensive ablation studies to evaluate the impact of various hyperparameters in the model.
format Preprint
id arxiv_https___arxiv_org_abs_2604_25925
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle SpecTr-GBV: Multi-Draft Block Verification Accelerating Speculative Decoding
Lin, Yijun
Sheng, Jinhao
Cai, Qingyue
Zhou, Feng
Computation and Language
Autoregressive language models suffer from high inference latency due to their sequential decoding nature. Speculative decoding (SD) mitigates this by employing a lightweight draft model to propose candidate tokens, which are selectively verified by a larger target model. While existing methods either adopt multi-draft strategies to increase acceptance rates or block verification techniques to jointly verify multiple tokens, they remain limited by treating these improvements in isolation. In this work, we propose SpecTr-GBV, a novel SD method that unifies multi-draft and greedy block verification (GBV) into a single framework. By formulating the verification step as an optimal transport problem over draft and target token blocks, SpecTr-GBV improves both theoretical efficiency and empirical performance. We theoretically prove that SpecTr-GBV achieves the optimal expected acceptance length physically attainable within the framework of i.i.d. draft generation, and this bound improves as the number of drafts increases. Empirically, we evaluate SpecTr-GBV across five datasets and four baselines. Our method achieves superior speedup and significantly higher block efficiency while preserving output quality. In addition, we perform comprehensive ablation studies to evaluate the impact of various hyperparameters in the model.
title SpecTr-GBV: Multi-Draft Block Verification Accelerating Speculative Decoding
topic Computation and Language
url https://arxiv.org/abs/2604.25925