paper-with-me

홈 › Papers

MetaLA: Unified Optimal Linear Approximation to Softmax Attention Map

2024-11-16 · Yuhong Chou, Man Yao, Kexin Wang, Yuqi Pan, Ruijie Zhu, Yiran Zhong, Yu Qiao, Jibin Wu, Bo Xu, Guoqi Li

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.

📄 PDF Abstract BibTeX arXiv:2411.10741

Code (1)

BICLab/MetaLA 공식 구현 pytorch

Tasks

image-classificationImage ClassificationLanguage ModelingLanguage Modelling

Methods 이 논문이 사용한 방법론

Attention 설명 없음
Adam 설명 없음
Residual Connection 설명 없음
BPE Byte Pair Encoding, or BPE, is a subword segmentation algorithm that encodes rare and unknown words as sequences of subword units. The intuition is that various word…
Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Absolute Position Encodings Absolute Position Encodings are a type of position embeddings for [Transformer-based models] where positional encodings are…
Multi-Head Attention 설명 없음
Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…

Similar Papers 제목 키워드 기반

Linear $Q$-Learning Does Not Diverge: Convergence Rates to a Bounded Set

2025-01-31 · Xinyu Liu, Zixuan Xie, Shangtong Zhang

$Q$-learning is one of the most fundamental reinforcement learning algorithms. Previously, it is widely believed that $Q$-learning with linear function approximation (i.e., linear $Q$-learning) suffers from possible dive…

Q-Learning

Rethinking the Global Convergence of Softmax Policy Gradient with Linear Function Approximation

2025-05-06 · Max Qiushi Lin, Jincheng Mei, Matin Aghaei, Michael Lu 외

Policy gradient (PG) methods have played an essential role in the empirical successes of reinforcement learning. In order to handle large state-action spaces, PG methods are typically used with function approximation. In…

Degrees of Freedom for Linear Attention: Distilling Softmax Attention with Optimal Feature Efficiency

2025-07-04 · Naoki Nishikawa, Rei Higuchi, Taiji Suzuki arxiv

Linear attention has attracted interest as a computationally efficient approximation to softmax attention, especially for long sequences. Recent studies have explored distilling softmax attention in pre-trained Transform…

Speaking of Language: Reflections on Metalanguage Research in NLP

2026-04-03 · Nathan Schneider, Antonios Anastasopoulos arxiv

This work aims to shine a spotlight on the topic of metalanguage. We first define metalanguage, link it to NLP and LLMs, and then discuss our two labs' metalanguage-centered efforts. Finally, we discuss four dimensions o…

Inference on Optimal Dynamic Policies via Softmax Approximation

2023-03-08 · Qizhao Chen, Morgane Austern, Vasilis Syrgkanis

Estimating optimal dynamic policies from offline data is a fundamental problem in dynamic decision making. In the context of causal inference, the problem is known as estimating the optimal dynamic treatment regime. Even…

Causal InferenceDecision Makingvalid