---
title: "QRAE: Query-Relevant Attention Entropy for Context Compression in DeepSearch Agents"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2611.07976v1"
permalink: "https://feed.astralor.com/items/itm_2652a01a66a535138ab2d837"
published: "2026-11-05T13:08:43.000Z"
score: 58
channel: "tech"
tags: ["LLM","AI Agent","推理效率","RAG","QRAE","Query-Relevant Attention Entropy","Context Compression","DeepSearch Agents DeepSearch"]
---

# QRAE: Query-Relevant Attention Entropy for Context Compression in DeepSearch Agents

## 速览

DeepSearch LLM agents repeatedly search and carry retrieved sources into subsequent reasoning steps. This expanding context increases token consumption and latency, although many results do not contribute to later reaso…

## 关键点

- DeepSearch LLM agents repeatedly search and carry retrieved sources into subsequent reasoning steps.
- This expanding context increases token consumption and latency, although many results do not contribute to later reasoning.
- We propose Query-Relevant Attention Entropy (QRAE), an online context compression method that uses sparse evidence-assessment heads to decide which intact retrieved chunks should persist across reasoning steps.

## 全文

DeepSearch LLM agents repeatedly search and carry retrieved sources into subsequent reasoning steps. This expanding context increases token consumption and latency, although many results do not contribute to later reasoning. We propose Query-Relevant Attention Entropy (QRAE), an online context compression method that uses sparse evidence-assessment heads to decide which intact retrieved chunks should persist across reasoning steps. It scores bounded candidate batches through paired prefills: a query-free control suppresses shared attention responses, while normalized entropy measures the breadth of the positive response within each chunk. High-scoring chunks enter a persistent evidence whitelist and are included in the reasoning context. Uncertain chunks remain outside the reasoning context but are retained in a candidate cache, where they can be retrieved for later reevaluation. QRAE requires neither an additional generative compression pass nor modification of the attention operator. Across multiple DeepSearch benchmarks and reasoning models, QRAE improves task success by an average of 2.35 percentage points and reduces end-to-end latency and reported token usage by an average of 30% and 45%, respectively. These benefits generalize to a model with nearly 100B parameters.

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_2652a01a66a535138ab2d837