paper-with-me

홈 › Papers

Multi-matrix Factorization Attention

2024-12-26 · Jingcheng Hu, Houyi Li, Yinmin Zhang, Zili Wang, Shuigeng Zhou, Xiangyu Zhang, Heung-Yeung Shum, Daxin Jiang

We propose novel attention architectures, Multi-matrix Factorization Attention (MFA) and MFA-Key-Reuse (MFA-KR). Existing variants for standard Multi-Head Attention (MHA), including SOTA methods like MLA, fail to maintain as strong performance under stringent Key-Value cache (KV cache) constraints. MFA enhances model capacity by efficiently scaling up both the number and dimension of attention heads through low-rank matrix factorization in the Query-Key (QK) circuit. Extending MFA, MFA-KR further reduces memory requirements by repurposing the key cache as value through value projection re-parameterization. MFA's design enables strong model capacity when working under tight KV cache budget, while MFA-KR is suitable for even harsher KV cache limits with minor performance trade-off. Notably, in our extensive and large-scale experiments, the proposed architecture outperforms MLA and performs comparably to MHA, while reducing KV cache usage by up to 56% and 93.7%, respectively.

📄 PDF Abstract BibTeX arXiv:2412.19255

Code (0)

등록된 구현이 없습니다.

Methods 이 논문이 사용한 방법론

Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…
Attention 설명 없음
Multi-Head Attention 설명 없음

Similar Papers 제목 키워드 기반

Multi-Mode Deep Matrix and Tensor Factorization

2021-09-29 · ICLR 2022 4 · Jicong Fan

Recently, deep linear and nonlinear matrix factorizations gain increasing attention in the area of machine learning. Existing deep nonlinear matrix factorization methods can only exploit partial nonlinearity of the data …

Tensor Decomposition

Graph-based Neural Acceleration for Nonnegative Matrix Factorization

2022-02-01 · Jens Sjölund, Maria Bånkestad

We describe a graph-based neural acceleration technique for nonnegative matrix factorization that builds upon a connection between matrices and bipartite graphs that is well-known in certain fields, e.g., sparse linear a…

Graph Neural Network

Multi-view Clustering via Unified Multi-kernel Learning and Matrix Factorization

2024-12-12 · Chenxing Jia, Mingjie Cai, Hamido Fujita

Multi-view clustering has become increasingly important due to the multi-source character of real-world data. Among existing multi-view clustering methods, multi-kernel clustering and matrix factorization-based multi-vie…

Clustering

Zipf Matrix Factorization : Matrix Factorization with Matthew Effect Reduction

2021-06-01 · Hao Wang

Recommender system recommends interesting items to users based on users' past information history. Researchers have been paying attention to improvement of algorithmic performance such as MAE and precision@K. Major techn…

FairnessLearning-To-RankRecommendation Systems

Interpretable Multi-View Clustering Based on Anchor Graph Tensor Factorization

2024-04-01 · Rui Wang, Jing Li, Quanxue Gao, Cheng Deng

The clustering method based on the anchor graph has gained significant attention due to its exceptional clustering performance and ability to process large-scale data. One common approach is to learn bipartite graphs wit…

Clustering