When Does Better Acceptance Pay? A Controlled Study of Speculative Decoding

Abstract

The throughput gains of speculative decoding depend on the trade-off between acceptance length and execution latency. We ask when additional acceptance is worth the cost of obtaining it. We establish a comparison protocol spanning training, inference, and evaluation, remeasuring released checkpoints and retraining six drafter recipes on shared target-generated data across ten benchmarks and two target models. Alignment changes the picture: gaps that appear large in published comparisons shrink substantially under shared training and inference conditions. This narrowing motivates two deeper questions: how is acceptance improved at different costs, and when do these gains improve throughput? We find that acceptance can be improved through nearly free objective and context changes or more expensive token-to-token conditioning. DCore illustrates the former route as a strong baseline with little additional measured drafting overhead. The payoff is conditional: task uncertainty and serving conditions affect acceptance returns, execution costs, and the preferred drafter and budget. These findings inform future drafter design and deployment choices.

Full paper on arXiv

Details

arXiv identifier

2610.24993

Published

2026-10-20

Authors

Yongchao Long, Chunrui Zheng, Yuxuan Zhou, Guangkun Nie, Tian-Chi Ma, Chao Pang, Shuohuan Wang, Yuchen Ding, Xianbin Wen, Dianhai Yu, Junyuan Shang, Yuxi Zhou, Shenda Hong

Open Models Built on This Paper

Every model in the SAVRN Model Hub whose card cites this paper, most downloaded first, with what it takes to run each one.