PSD: Pushing the Pareto Frontier of Diffusion LLMs via Parallel Speculative Decoding
arXiv cs.CL·Shengyin Sun, Yiming Li, Renxi Liu, Xinqi Li, Hui-Ling Zhen, Weizhe Lin, Chen Chen, Xianzhi Yu, Mingxuan Yuan, Chen Ma
4d agoQuick Take
PSD enhances diffusion LLMs' inference efficiency and quality using adaptive unmasking and speculative drafts.
Key Points
- Introduces Parallel Speculative Decoding (PSD) framework.
- Reduces inference cost via adaptive unmasking policy.
- Achieves up to 5.5x tokens per forward pass.
📖 Reader Mode
~2 min readAbstract:Diffusion large language models (dLLMs) generate text by iteratively denoising masked token sequences. Although dLLMs can predict all masked positions in parallel within each step, the large number of denoising iterations still makes inference expensive. This cost can be reduced spatially by unmasking multiple tokens per step, or temporally by collapsing multiple denoising steps into one verification call. We propose Parallel Speculative Decoding (PSD), a training-free framework that jointly improves inference along both axes. Using the confidence scores from a single forward pass, PSD selects positions to unmask via a configurable, adaptive unmasking policy and constructs multi-depth speculative drafts without extra model calls. A final batched verification pass then applies hierarchical acceptance, keeping the deepest draft that remains consistent with the updated predictions. Experiments on three dLLMs across reasoning and code generation tasks show that PSD achieves favorable trade-offs between inference efficiency and generation quality, reaching up to $5.5\times$ tokens per forward pass with accuracy comparable to greedy decoding.
| Comments: | 16 pages |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2605.15609 [cs.CL] |
| (or arXiv:2605.15609v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2605.15609 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Shengyin Sun [view email]
[v1]
Fri, 15 May 2026 04:43:02 UTC (300 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CL
See more →arXiv cs.CL·Leyao Wang, Yanan He, Peng Chen, Asaf Yehudai, Yixin Liu, Rex Ying, Michal Shmueli-Scheuer, Arman Cohan
2d ago FeaturedOriginal
Time to REFLECT: Can We Trust LLM Judges for Evidence-based Research Agents?
AI Summary
The reliability of LLM judges for evaluating deep research agents is critically assessed using the REFLECT benchmark.