English(EN) hLLM: Single Pass Decoding for Generative Reranking

hLLM解码策略将LLM重排速度提升64倍

作者 PulseAugur 编辑部 · [2 个来源] · 2026-09-01 19:30

研究人员开发了hLLM，一种用于大型语言模型（LLM）的新型解码策略，可显著加快生成式重排的速度。通过将输出视为一个可用匈牙利算法解决的排列问题，hLLM实现了$64 imes$的端到端推理加速，将其缩短至28毫秒。该方法保持了与现有教师模型相当的排名质量，并将生成式排名与组合优化联系起来。 AI

影响这种新的解码策略可以为基于LLM的排名系统实现实时应用。

排序理由该集群包含一篇详细介绍LLM解码新方法的论文。

AI 生成摘要 · Google Gemini · 来自 2 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 1 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Research; 该集群包含一篇详细介绍LLM解码新方法的论文。
Source corroboration: 2 independent sources; Multiple independent publishers reporting the same story raises confidence that it's real and newsworthy.
Topics: paper, infra; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: 1 days old; Coverage has settled into its steady-state source set.

完整方法见我们的编辑标准。

arXiv cs.AI TIER_1 English(EN) · Emil Laftchiev, Prachi Agrawal, Moe Kayali, Bixing Yan, Qi Xu, Zijie Lei, Chen Qiu, Zhi Hua, Ke Li, Luke Simon · 2026-09-03 04:00

hLLM：生成式重排的单次解码

arXiv:2609.01807v1 Announce Type: cross Abstract: Large language models (LLMs) achieve state-of-the-art generative ranking quality, but the ranking they produce must be decoded, and autoregressive decoding spends one sequential forward pass per emitted token. We observe that the …
arXiv cs.IR (Information Retrieval) TIER_1 English(EN) · Luke Simon · 2026-09-01 19:30

hLLM：生成式重排的单次解码

Large language models (LLMs) achieve state-of-the-art generative ranking quality, but the ranking they produce must be decoded, and autoregressive decoding spends one sequential forward pass per emitted token. We observe that the only tokens a ranker must emit are the $N$ ordinal…