DCOCO：面向推测解码的并行草稿上下文校正
原标题：Contextual Correction of Parallel Draftsfor Speculative Decoding
AI 摘要
该论文提出 DCOCO（DFlash Contextual Corrector），一种用于并行扩散草稿器（如 DFlash）的轻量级上下文校正机制，旨在恢复推测解码中因并行生成而丢失的块内依赖关系。DCOCO 不直接采用每个位置的 top-1 预测，而是保留多个候选 token，并利用草稿器已有的隐藏状态和 logits 并行评估相邻位置间的上下文兼容性，再通过前缀优先的贪心选择规则解码。在四个验证器骨干和十个基准上，DCOCO 在所有配置中取得最高平均吞吐量，相比原始解码最高加速 6.6 倍；在 4–32 并发请求下比最强竞争方法高 6–23%。
正文节选
Contextual Correction of Parallel Drafts for Speculative Decoding Abstract 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 autoregressiv