提出解耦对比解码，提升生成速度且不牺牲质量。
Decoupled Contrastive Decoding via Expert-Aligned Drafting
- 用专家对齐的轻量级草案器生成候选，仅在验证阶段使用对比信号。
- 相比原版对比解码提速1.65至1.95倍，提案路径延迟降低5到12倍。
- 适合追求高效高质生成的模型部署与推理优化场景。
对比解码（CD）虽能提升生成质量，但其依赖业余模型进行草案生成，导致解码成本高昂。若通过推测解码加速CD，需回答一个核心问题：对比信号应作用于草案生成，还是仅用于验证？本文在轻量级特征级草案器框架下研究此问题。通过匹配的Cross-alpha训练与近似双草案器分解两种控制性诊断，结果一致表明：具备对比感知的草案生成并未持续优于专家对齐草案生成，因为对比修正通常弱于草案误差，且重建过程可能放大该误差。为此，我们提出解耦对比解码（DCD），即使用专家对齐的轻量级提案器生成草案，并仅在不变的对比解码验证阶段引入业余模型。标准推测验证保留了原始CD的输出分布。在主流8B设置下，基于EAGLE3的DCD相较原版CD平均获得1.65至1.95倍的贪婪加速，并使MMLU提案路径延迟降低约5至12倍。
原文摘要 · Abstract (English)
Contrastive Decoding (CD) improves generation quality, but its amateur-model pass makes decoding expensive. Accelerating CD with speculative decoding raises a proposal-alignment question: should the contrastive signal shape the drafter, or should it remain only in verification? We study this question in the lightweight feature-level drafter regime. Two controlled diagnostics, matched Cross-alpha training and an Approximate Dual-Drafter decomposition, give the same diagnosis: contrastive-aware drafting does not consistently improve over expert-aligned drafting because the contrastive correction is usually weaker than drafter error, and reconstruction can amplify that error. We introduce Decoupled Contrastive Decoding (DCD), which drafts with an expert-aligned lightweight proposer and applies the amateur only in unchanged CD verification. Standard speculative verification preserves the vanilla-CD output distribution. Across the main 8B settings, EAGLE3-based DCD achieves average greedy speedups of 1.65 to 1.95x over vanilla CD and reduces MMLU proposal-path latency by about 5 to 12x relative to amateur-coupled proposal paths.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。