---
id: 20261017-T0-06
title: "新方法用泰勒级数将Transformer注意力降到线性时间"
title_en: "Taylor-Series Approach Achieves Linear-Time Transformer Attention with Fixed Recurrent States"
url: https://ai.daily.yangsir.net/daily/20261017-T0-06
issue_date: 2026-10-17
publish_date: 2026-10-16T04:00:00.000Z
category: research
source_name: "arXiv cs.LG (ML)"
source_url: https://arxiv.org/abs/2610.20637
---

# 新方法用泰勒级数将Transformer注意力降到线性时间

研究提出一种将注意力指数核展开为n项截断泰勒级数的方法，每项分解为张量幂内积，从而将因果历史压缩为恰好n个循环状态，每个状态通过一次秩一外积更新。该方法在保持固定数量循环状态的前提下，将Transformer注意力计算复杂度降至线性时间。这为解决长序列场景下注意力机制的二次复杂度瓶颈提供了新思路。

## English Version

**Taylor-Series Approach Achieves Linear-Time Transformer Attention with Fixed Recurrent States**

Researchers propose expanding the exponential kernel of attention into a truncated Taylor series of n terms, factorizing each into a tensor-power inner product. This collapses causal history into exactly n recurrent states, each updated by a rank-one outer product. The method achieves linear-time attention computation with a fixed number of recurrent states, offering a new approach to the quadratic complexity bottleneck in long-sequence Transformer attention.

---

**来源**：[arXiv cs.LG (ML)](https://arxiv.org/abs/2610.20637)

**详情页**：https://ai.daily.yangsir.net/daily/20261017-T0-06

---

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