Contextual Correction of Parallel Draftsfor Speculative Decoding
Quick Answer
This paper shows that DCOCO (DFlash Contextual Corrector) enhances speculative decoding by restoring intra-block dependencies while maintaining parallel token drafting, achieving up to 6.6x speedup over traditional methods.
Quick Take
It outperforms competitors by 6-23% across 4-32 concurrent requests, demonstrating consistent acceleration in various decoding regimes and loads.
Key Points
- DCOCO retains multiple candidate tokens for contextual compatibility evaluation.
- Achieves highest average throughput across four verifier backbones and ten benchmarks.
- Demonstrates speedups of up to 6.6x over vanilla decoding methods.
- Outperforms competing methods by 6-23% under concurrent serving conditions.
- Restores intra-block dependencies without sacrificing parallelism.
Paper Resources
📖 Reader Mode
~2 min readAbstract:Speculative decoding accelerates large language model inference by using a lightweight drafter to propose multiple tokens that are subsequently verified in parallel by the target model. Recent diffusion-based drafters make this paradigm particularly attractive: deep attention-based neural backbones provide strong predictions, while drafting multiple future positions in parallel avoids the sequential cost of autoregressive generation. This parallelism, however, comes at a cost. Tokens within a draft block are predicted without access to the realized identities of preceding draft tokens, removing intra-block dependencies that would naturally be available under autoregressive generation. We introduce DCOCO (DFlash Contextual Corrector), a lightweight mechanism that restores these dependencies while preserving parallel drafting. Rather than committing to each position's top prediction, DCOCO retains multiple candidate tokens and evaluates their contextual compatibility across adjacent draft positions in parallel, enabling contextually informed token selection at low additional cost. Across four verifier backbones and ten benchmarks, DCOCO achieves the highest average throughput in every evaluated configuration, spanning greedy and stochastic decoding with and without extended reasoning, with speedups of up to 6.6x over vanilla decoding. Its advantage persists under concurrent serving: DCOCO outperforms the strongest competing method by 6-23% across 4-32 concurrent requests, while other sequential correctors fall below uncorrected DFlash at these loads. These results demonstrate that recovering intra-block dependencies through parallel contextual scoring delivers consistent acceleration across decoding regimes and serving loads.
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.12570 [cs.CL] |
| (or arXiv:2610.12570v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.12570 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Roy Eisenstadt [view email]
[v1]
Thu, 8 Oct 2026 18:01:03 UTC (236 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 →The "10th Juror": Open-Set Standpoint Screening for Bureaucratic Bias Detection
MARS-Gov introduces a framework for detecting bureaucratic bias in Dutch government documents, achieving a new state-of-the-art F1 score of 0.880. This model outperforms existing zero-shot detectors by 20.2 points and reduces unnecessary interventions to just 2.5%. The framework's dynamic '10th juror' adapts to emerging biases, enhancing legal language processing.