English
Related papers

Related papers: Routing Absorption in Sparse Attention: Why Random…

200 papers

Adaptive control of Euler-Lagrange systems is challenging when friction is governed by a finite-horizon internal state that is not directly observable from joint measurements. In this setting, the measured closed-loop state is no longer…

Machine Learning · Computer Science 2026-05-11 Giansalvo Cirrincione , Adriano Fagiolini

Transformers have been successfully applied to the visual tracking task and significantly promote tracking performance. The self-attention mechanism designed to model long-range dependencies is the key to the success of Transformers.…

Computer Vision and Pattern Recognition · Computer Science 2022-05-10 Zhihong Fu , Zehua Fu , Qingjie Liu , Wenrui Cai , Yunhong Wang

Large language models can follow complex instructions in a single turn, yet over long multi-turn interactions they often lose the thread of instructions, persona, and rules. This degradation has been measured behaviorally but not…

Artificial Intelligence · Computer Science 2026-05-14 Vardhan Dongre , Joseph Hsieh , Viet Dac Lai , Seunghyun Yoon , Trung Bui , Dilek Hakkani-Tür

Transformers are state-of-the-art models for a variety of sequence modeling tasks. At their core is an attention function which models pairwise interactions between the inputs at every timestep. While attention is powerful, it does not…

Computation and Language · Computer Science 2021-03-23 Hao Peng , Nikolaos Pappas , Dani Yogatama , Roy Schwartz , Noah A. Smith , Lingpeng Kong

Transformer models have been widely adopted in various domains over the last years, and especially large language models have advanced the field of AI significantly. Due to their size, the capability of these networks has increased…

Machine Learning · Computer Science 2023-11-10 Yelysei Bondarenko , Markus Nagel , Tijmen Blankevoort

The quadratic complexity of attention mechanisms poses a critical bottleneck for large language models processing long contexts. While dynamic sparse attention methods offer input-adaptive efficiency, they face fundamental trade-offs:…

Computation and Language · Computer Science 2026-02-06 Siran Liu , Guoxia Wang , Sa Wang , Jinle Zeng , HaoYang Xie , Siyu Lou , JiaBin Yang , DianHai Yu , Haifeng Wang , Chao Yang

Transformer architecture has been very successful long runner in the field of Deep Learning (DL) and Large Language Models (LLM) because of its powerful attention-based learning and parallel-natured architecture. As the models grow gigantic…

Machine Learning · Computer Science 2026-01-21 Phani Kumar , Nyshadham , Jyothendra Varma , Polisetty V R K , Aditya Rathore

To perform few-shot learning, language models extract signals from a few input-label pairs, aggregate these into a learned prediction rule, and apply this rule to new inputs. How is this implemented in the forward pass of modern transformer…

Machine Learning · Computer Science 2025-10-10 Xinyan Hu , Kayo Yin , Michael I. Jordan , Jacob Steinhardt , Lijie Chen

The computational demands of self-attention mechanisms pose a critical challenge for transformer-based video generation, particularly in synthesizing ultra-long sequences. Current approaches, such as factorized attention and fixed sparse…

Computer Vision and Pattern Recognition · Computer Science 2025-08-19 Qirui Li , Guangcong Zheng , Qi Zhao , Jie Li , Bin Dong , Yiwu Yao , Xi Li

Graph Transformers (GTs) have demonstrated superior performance compared to traditional message-passing graph neural networks in many studies, especially in processing graph data with long-range dependencies. However, GTs tend to suffer…

Machine Learning · Computer Science 2025-04-30 Zhonghao Li , Ji Shi , Xinming Zhang , Miao Zhang , Bo Li

Scaling large language models to long contexts is challenging due to the quadratic computational cost of full attention. Mitigation approaches include KV-cache selection or compression techniques. We instead provide an effective and…

Machine Learning · Computer Science 2026-04-24 Yuzhen Mao , Michael Y. Li , Emily B. Fox

With the attention mechanism, transformers achieve significant empirical successes. Despite the intuitive understanding that transformers perform relational inference over long sequences to produce desirable representations, we lack a…

Machine Learning · Computer Science 2024-04-02 Yufeng Zhang , Boyi Liu , Qi Cai , Lingxiao Wang , Zhaoran Wang

The transformer architecture has been widely applied to many machine learning tasks. A main bottleneck in the time to perform transformer computations is a task called attention computation. [Alman and Song, NeurIPS 2023] have shown that in…

Machine Learning · Computer Science 2025-05-20 Josh Alman , Zhao Song

We present the Condensate Theorem: attention sparsity is a learned topological property, not an architectural constraint. Through empirical analysis of trained language models, we find that attention mass concentrates on a distinct…

Machine Learning · Computer Science 2026-02-11 Jorge L. Ruiz Williams

The sparse transformer can reduce the computational complexity of the self-attention layers to $O(n)$, whilst still being a universal approximator of continuous sequence-to-sequence functions. However, this permutation variant operation is…

Machine Learning · Computer Science 2023-03-01 Shidi Li , Christian Walder , Alexander Soen , Lexing Xie , Miaomiao Liu

The interpretability of Mixture-of-Experts (MoE) models, especially those with heterogeneous designs, remains underexplored. Existing attribution methods for dense models fail to capture dynamic routing-expert interactions in sparse MoE…

Computation and Language · Computer Science 2025-06-12 Junzhuo Li , Bo Wang , Xiuze Zhou , Peijie Jiang , Jia Liu , Xuming Hu

Mechanistic interpretability improves the safety, reliability, and robustness of large AI models. This study examined individual attention heads in vision transformers (ViTs) fine tuned on distorted 2D spectrogram images containing non…

Machine Learning · Computer Science 2025-03-25 Nooshin Bahador

In the Transformer model, "self-attention" combines information from attended embeddings into the representation of the focal embedding in the next layer. Thus, across layers of the Transformer, information originating from different tokens…

Machine Learning · Computer Science 2020-06-02 Samira Abnar , Willem Zuidema

In mechanistic interpretability, recent work scrutinizes transformer "circuits" - sparse, mono or multi layer sub computations, that may reflect human understandable functions. Yet, these network circuits are rarely acid-tested for their…

Machine Learning · Computer Science 2026-02-20 Karan Bali , Jack Stanley , Praneet Suresh , Danilo Bzdok

Several recent Transformer architectures expose later layers to representations computed in the earliest layers, motivated by the observation that low-level features can become harder to recover as the residual stream is repeatedly…

Machine Learning · Computer Science 2026-05-07 Skye Gunasekaran , Téa Wright , Rui-Jie Zhu , Jason Eshraghian