English
Related papers

Related papers: MDN: Parallelizing Stepwise Momentum for Delta Lin…

200 papers

Linear attention and state-space models offer constant-memory alternatives to softmax attention, but often struggle with in-context associative recall. The Delta Rule mitigates this by writing each token via one step of online gradient…

Machine Learning · Computer Science 2026-05-14 Chenyu Zhou , Hongpei Li , Yuerou Liu , Jianghao Lin , Dongdong Ge , Yinyu Ye

We present Lightning Attention, the first linear attention implementation that maintains a constant training speed for various sequence lengths under fixed memory consumption. Due to the issue with cumulative summation operations (cumsum),…

Computation and Language · Computer Science 2024-06-21 Zhen Qin , Weigao Sun , Dong Li , Xuyang Shen , Weixuan Sun , Yiran Zhong

Linear Recurrent Neural Networks (linear RNNs) have emerged as competitive alternatives to Transformers for sequence modeling, offering efficient training and linear-time inference. However, existing architectures face a fundamental…

Machine Learning · Computer Science 2025-10-24 Julien Siems , Timur Carstensen , Arber Zela , Frank Hutter , Massimiliano Pontil , Riccardo Grazzi

Recent efficient sequence modeling methods such as Gated DeltaNet, TTT, and RWKV-7 have achieved performance improvements by supervising the recurrent memory management through Delta learning rule. Unlike previous state-space models (e.g.,…

Machine Learning · Computer Science 2025-12-04 Jiaxi Hu , Yongqi Pan , Jusen Du , Disen Lan , Xiaqiang Tang , Qingsong Wen , Yuxuan Liang , Weigao Sun

Transformer-based large language models (LLMs) have achieved remarkable success, yet their standard attention mechanism incurs quadratic computation and memory costs with respect to sequence length, posing a major bottleneck for…

Machine Learning · Computer Science 2025-10-22 Tao Bu , Qiangang Wang , Bowen Zeng , Hanwen Sun , Yunpeng Huang , Chun Cao , Jingwei Xu

Deploying Large Language Models (LLMs) on edge devices remains challenging due to their quadratically increasing computations with the sequence length. Existing studies for dynamic attention pruning are designed for hardware with massively…

Artificial Intelligence · Computer Science 2025-07-29 Jiawen Qi , Chang Gao , Zhaochun Ren , Qinyu Chen

Sequence modeling is currently dominated by causal transformer architectures that use softmax self-attention. Although widely adopted, transformers require scaling memory and compute linearly during inference. A recent stream of work…

Large language models (LLMs) with billions of parameters demonstrate impressive performance. However, the widely used Multi-Head Attention (MHA) in LLMs incurs substantial computational and memory costs during inference. While some efforts…

Machine Learning · Computer Science 2024-12-10 Yilong Chen , Linhao Zhang , Junyuan Shang , Zhenyu Zhang , Tingwen Liu , Shuohuan Wang , Yu Sun

Transformers are highly parallel but are limited to computations in the TC$^0$ complexity class, excluding tasks such as entity tracking and code execution that provably require greater expressive power. Motivated by this limitation, we…

Machine Learning · Computer Science 2026-05-15 Mayank Mishra , Shawn Tan , Ion Stoica , Joseph Gonzalez , Tri Dao

Linear attention methods offer a compelling alternative to softmax attention due to their efficiency in recurrent decoding. Recent research has focused on enhancing standard linear attention by incorporating gating while retaining its…

Machine Learning · Computer Science 2025-04-08 Yingcong Li , Davoud Ataee Tarzanagh , Ankit Singh Rawat , Maryam Fazel , Samet Oymak

Linear transformers have emerged as a subquadratic-time alternative to softmax attention and have garnered significant interest due to their fixed-size recurrent state that lowers inference cost. However, their original formulation suffers…

Computation and Language · Computer Science 2024-05-13 Jean Mercat , Igor Vasiljevic , Sedrick Keh , Kushal Arora , Achal Dave , Adrien Gaidon , Thomas Kollar

Foundation models, now powering most of the exciting applications in deep learning, are almost universally based on the Transformer architecture and its core attention module. Many subquadratic-time architectures such as linear attention,…

Machine Learning · Computer Science 2024-06-03 Albert Gu , Tri Dao

Large Language Models (LLMs) have become the central paradigm in artificial intelligence, yet the core computational primitive of attention has remained structurally unchanged. Local Linear Attention (LLA) is an attention mechanism derived…

Machine Learning · Computer Science 2026-05-29 Yifei Zuo , Dhruv Pai , Zhichen Zeng , Alec Dewulf , Shuming Hu , Zhaoran Wang

The original softmax-based attention mechanism (regular attention) in the extremely successful Transformer architecture computes attention between $N$ tokens, each embedded in a $D$-dimensional head, with a time complexity of $O(N^2D)$.…

Machine Learning · Computer Science 2025-10-28 Armin Gerami , Ramani Duraiswami

Efficiently training LLMs with long sequences is important yet challenged by the massive computation and memory requirements. Sequence parallelism has been proposed to tackle these problems, but existing methods suffer from scalability or…

Distributed, Parallel, and Cluster Computing · Computer Science 2024-06-27 Diandian Gu , Peng Sun , Qinghao Hu , Ting Huang , Xun Chen , Yingtong Xiong , Guoteng Wang , Qiaoling Chen , Shangchun Zhao , Jiarui Fang , Yonggang Wen , Tianwei Zhang , Xin Jin , Xuanzhe Liu

While the Transformer architecture dominates many fields, its quadratic self-attention complexity hinders its use in large-scale applications. Linear attention offers an efficient alternative, but its direct application often degrades…

Computer Vision and Pattern Recognition · Computer Science 2026-01-15 Kewei Zhang , Ye Huang , Yufan Deng , Jincheng Yu , Junsong Chen , Huan Ling , Enze Xie , Daquan Zhou

Transformer architectures have achieved remarkable success in various domains. While efficient alternatives to Softmax Attention have been widely studied, the search for more expressive mechanisms grounded in theoretical insight-even at…

Machine Learning · Computer Science 2025-10-03 Yifei Zuo , Yutong Yin , Zhichen Zeng , Ang Li , Banghua Zhu , Zhaoran Wang

Transformer-based models have emerged as a leading architecture for natural language processing, natural language generation, and image generation tasks. A fundamental element of the transformer architecture is self-attention, which allows…

Machine Learning · Computer Science 2025-07-01 Venmugil Elango

The quadratic computational complexity of MultiHead SelfAttention (MHSA) remains a fundamental bottleneck in scaling Large Language Models (LLMs) for longcontext tasks. While sparse and linearized attention mechanisms attempt to mitigate…

Computation and Language · Computer Science 2025-12-19 Caner Erden

MEGA is a recent transformer-based architecture, which utilizes a linear recurrent operator whose parallel computation, based on the FFT, scales as $O(LlogL)$, with $L$ being the sequence length. We build upon their approach by replacing…

Machine Learning · Computer Science 2023-12-12 Aleksandar Terzic , Michael Hersche , Geethan Karunaratne , Luca Benini , Abu Sebastian , Abbas Rahimi