paper-with-me

Papers

Variational Speculative Decoding: Rethinking Draft Training from Token Likelihood to Sequence Acceptance

2026-02-05 · Xiandong Zou, Jianshu Li, Jing Huang, Pan Zhou arxiv

Speculative decoding accelerates inference for (M)LLMs, yet a training-decoding discrepancy persists: while existing methods optimize single greedy trajectories, decoding involves verifying and ranking multiple sampled draft paths. We propose Variational Speculative Decoding (VSD), formulating draft training as variational inference over latent proposals (draft paths). VSD maximizes the marginal probability of target-model acceptance, yielding an ELBO that promotes high-quality latent proposals while minimizing divergence from the target distribution. To enhance quality and reduce variance, we incorporate a path-level utility and optimize via an Expectation-Maximization procedure. The E-step draws Monte Carlo samples from an oracle-filtered posterior, while the M-step maximizes weighted likelihood using Adaptive Rejection Weighting (ARW) and Confidence-Aware Regularization (CAR). Theoretical analysis confirms that VSD increases expected acceptance length and speedup. Extensive experiments across LLMs and MLLMs show that VSD achieves up to a 9.6% speedup over EAGLE-3 and 7.9% over ViSpec, significantly improving decoding efficiency.

📄 PDF Abstract BibTeX arXiv:2602.05774

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Fail Fast, Win Big: Rethinking the Drafting Strategy in Speculative Decoding via Diffusion LLMs

2025-12-23 · Rui Pan, Zhuofu Chen, Hongyi Liu, Arvind Krishnamurthy 외 arxiv

Diffusion Large Language Models (dLLMs) offer fast, parallel token generation, but their standalone use is plagued by an inherent efficiency-quality tradeoff. We show that, if carefully applied, the attributes of dLLMs c…

Decoding Speculative Decoding

2024-02-02 · Minghao Yan, Saurabh Agarwal, Shivaram Venkataraman

Speculative Decoding is a widely used technique to speed up inference for Large Language Models (LLMs) without sacrificing quality. When performing inference, speculative decoding uses a smaller draft model to generate s…

Language ModelingLanguage Modelling

SpecForge: A Flexible and Efficient Open-Source Training Framework for Speculative Decoding

2026-03-19 · Shenggui Li, Chao Wang, Yikai Zhu, Yubo Wang 외 arxiv

Large language models incur high inference latency due to sequential autoregressive decoding. Speculative decoding alleviates this bottleneck by using a lightweight draft model to propose multiple tokens for batched veri…

Ouroboros: Generating Longer Drafts Phrase by Phrase for Faster Speculative Decoding

2024-02-21 · Weilin Zhao, Yuxiang Huang, Xu Han, Wang Xu 외

Speculative decoding is a widely used method that accelerates the generation process of large language models (LLMs) with no compromise in model performance. It achieves this goal by using an existing smaller model for d…

Text Generation

ParallelSpec: Parallel Drafter for Efficient Speculative Decoding

2024-10-08 · Zilin Xiao, Hongming Zhang, Tao Ge, Siru Ouyang 외

Speculative decoding has proven to be an efficient solution to large language model (LLM) inference, where the small drafter predicts future tokens at a low cost, and the target model is leveraged to verify them in paral…

Language ModelingLanguage ModellingLarge Language ModelText Generation