# MetaLA: Unified Optimal Linear Approximation to Softmax Attention Map

Source: [MetaLA: Unified Optimal Linear Approximation to Softmax Attention Map](<https://www.sota2.com/research/paper/68fa75c41e209131cea69d2f>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-26T03:41:42.944Z

Authors: Yuhong Chou, Man Yao, Kexin Wang, Yuqi Pan, Ruijie Zhu, Yiran Zhong, Yu Qiao, Jibin Wu, Bo Xu, Guoqi Li

Published: 2024-11-16T08:47:32.000Z

## Abstract

Various linear complexity models, such as Linear Transformer (LinFormer), State Space Model (SSM), and Linear RNN (LinRNN), have been proposed to replace the conventional softmax attention in Transformer structures. However, the optimal design of these linear models is still an open question. In this work, we attempt to answer this question by finding the best linear approximation to softmax attention from a theoretical perspective. We start by unifying existing linear complexity models as the linear attention form and then identify three conditions for the optimal linear attention design: 1) Dynamic memory ability; 2) Static approximation ability; 3) Least parameter approximation. We find that none of the current linear models meet all three conditions, resulting in suboptimal performance. Instead, we propose Meta Linear Attention (MetaLA) as a solution that satisfies these conditions. Our experiments on Multi-Query Associative Recall (MQAR) task, language modeling, image classification, and Long-Range Arena (LRA) benchmark demonstrate that MetaLA is more effective than the existing linear models.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Image Classification on ImageNet-1K 1.0 (val)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-1-0-val.md>): Image Classification; ImageNet-1K 1.0 (val); Top-1 Accuracy: 75.3.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2411.10741>)

[PDF](<https://arxiv.org/pdf/2411.10741.pdf>)
