---
id: 20260521-T0-14
title: "D-PACE：提升LLM推理速度的并行解码技术"
title_en: "D-PACE: New Parallel Speculative Decoding for Faster LLM Inference"
url: https://ai.daily.yangsir.net/daily/20260521-T0-14
issue_date: 2026-05-21
publish_date: 2026-05-20T04:00:00.000Z
category: research
source_name: "arXiv cs.LG (ML)"
source_url: https://arxiv.org/abs/2605.18810
---

# D-PACE：提升LLM推理速度的并行解码技术

研究人员开发D-PACE技术，通过动态位置感知交叉熵优化并行解码过程。该方法将大模型验证阶段的计算量减少35%，同时保持输出质量不变。相比现有方案，D-PACE能够更准确预测下一个token块，显著降低了验证阶段的计算开销。该技术特别适合需要快速响应的LLM应用场景，如实时翻译和代码生成。论文已在arXiv发布。

## English Version

**D-PACE: New Parallel Speculative Decoding for Faster LLM Inference**

Researchers developed D-PACE, a new parallel speculative decoding technique using dynamic position-aware cross-entropy. It reduces verification computation by 35% while maintaining output quality. The method more accurately predicts next token blocks, significantly cutting verification overhead. Suitable for real-time applications like translation and code generation, the paper is available on arXiv.

---

**来源**：[arXiv cs.LG (ML)](https://arxiv.org/abs/2605.18810)

**详情页**：https://ai.daily.yangsir.net/daily/20260521-T0-14

---

*智语观潮 · Daily — https://ai.daily.yangsir.net/llms.txt*