---
id: 20261027-T0-07
title: "SPECMEM：用实时记忆管理加速Agent工具调用"
title_en: "SPECMEM Speeds Up Agent Tool Calling with Live Memory Management"
url: https://ai.daily.yangsir.net/daily/20261027-T0-07
issue_date: 2026-10-27
publish_date: 2026-10-26T04:00:00.000Z
category: research
source_name: "arXiv cs.CL (NLP)"
source_url: https://arxiv.org/abs/2610.31840
---

# SPECMEM：用实时记忆管理加速Agent工具调用

工具调用型Agent通常以准确率为优化目标，但落地时真正的成本在延迟：每次工具调用都要自回归解码。论文提出SPECMEM，用推测解码在不改变输出的前提下加速，并通过检索历史调用来提升草稿质量。这套思路把“记忆”放进解码环节，而不是只用于上下文拼接。对部署高频工具调用Agent的团队，延迟是直接影响用户体验和成本的关键指标。

## English Version

**SPECMEM Speeds Up Agent Tool Calling with Live Memory Management**

Tool-calling agents are usually optimized for accuracy, but in deployment their real cost is latency: every tool call is decoded autoregressively. The paper introduces SPECMEM, which applies speculative decoding to speed this up without changing outputs, and retrieves past calls to improve draft quality. The approach puts memory inside the decoding step rather than only in context assembly. For teams running high-frequency tool-calling agents, latency is a key metric affecting both user experience and cost.

---

**来源**：[arXiv cs.CL (NLP)](https://arxiv.org/abs/2610.31840)

**详情页**：https://ai.daily.yangsir.net/daily/20261027-T0-07

---

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