English(EN) Compression Beyond the Uncompressed: A Two-Stage Training Recipe for Soft Context Compression in RAG

新的RAG方法DEX-Comp实现16倍上下文压缩，4-24倍加速

作者 PulseAugur 编辑部 · [1 个来源] · 2026-09-07 04:00

研究人员开发了DEX-Comp，一种新颖的两阶段训练方法，以提高检索增强生成（RAG）系统的效率。该方法首先使用纯蒸馏在未压缩RAG的正确响应上训练压缩模型，然后采用强化学习的硬探索来处理未压缩系统失败的查询。DEX-Comp在各种基准测试中实现了16倍的压缩和4倍-24倍的推理加速，其性能与未压缩的RAG相当或更优。 AI

影响这项研究可以显著提高RAG系统的效率并降低计算成本，使其更适合广泛部署。

排序理由该集群包含一篇详细介绍改进RAG系统新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.CL 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 32 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该集群包含一篇详细介绍改进RAG系统新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
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: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.CL TIER_1 English(EN) · Shuyu Guo, Shuo Zhang, Zhaochun Ren · 2026-09-07 04:00

超越未压缩的压缩：用于RAG软上下文压缩的两阶段训练方法

arXiv:2609.05152v1 Announce Type: new Abstract: Retrieval-Augmented Generation (RAG) enhances language models with external knowledge, but the lengthy retrieved context inflates the input and degrades inference efficiency. Soft context compression encodes each document into a sub…

本文如何被排名

报道来源 [1]

超越未压缩的压缩：用于RAG软上下文压缩的两阶段训练方法

相关实体

相关话题