English
Related papers

Related papers: Attention Sinks Induce Gradient Sinks: Massive Act…

200 papers

We study two recurring phenomena in Transformer language models: massive activations, in which a small number of tokens exhibit extreme outliers in a few channels, and attention sinks, in which certain tokens attract disproportionate…

Artificial Intelligence · Computer Science 2026-03-06 Shangwen Sun , Alfredo Canziani , Yann LeCun , Jiachen Zhu

We investigate the functional role of emergent outliers in large language models, specifically attention sinks (a few tokens that consistently receive large attention logits) and residual sinks (a few fixed dimensions with persistently…

Despite the prevalence of the attention sink phenomenon in Large Language Models (LLMs), where initial tokens disproportionately monopolize attention scores, its structural origins remain elusive. This work provides a \textit{mechanistic…

Machine Learning · Computer Science 2026-05-08 Siquan Li , Kaiqi Jiang , Jiacheng Sun , Tianyang Hu

Attention sinks -- tokens that receive disproportionate attention mass -- are assumed to be functionally important in autoregressive language models, but their role in diffusion transformers remains unclear. We present a causal analysis in…

Computer Vision and Pattern Recognition · Computer Science 2026-05-13 Fangzheng Wu , Brian Summa

Attention sinks and compression valleys have attracted significant attention as two puzzling phenomena in large language models, but have been studied in isolation. In this work, we present a surprising connection between attention sinks…

Transformers commonly exhibit an attention sink: disproportionately high attention to the first position. We study this behavior in GPT-2-style models with learned query biases and absolute positional embeddings. Combining structural…

Machine Learning · Computer Science 2026-04-17 Yuval Ran-Milo , Hila Ofek , Shahar Mendel

Attention sinks are defined as tokens that attract disproportionate attention. While these have been studied in single modality transformers, their cross-modal impact in Large Vision-Language Models (LVLM) remains largely unexplored: are…

Computer Vision and Pattern Recognition · Computer Science 2026-04-07 Jiho Choi , Jaemin Kim , Sanghwan Kim , Seunghoon Hong , Jin-Hwi Park

Transformers often display an attention sink: probability mass concentrates on a fixed, content-agnostic position. Are sinks a byproduct of the optimization/training regime? Or are they sometimes functionally necessary in softmax…

Machine Learning · Computer Science 2026-04-20 Yuval Ran-Milo

Transformers empirically perform precise probabilistic reasoning in carefully constructed ``Bayesian wind tunnels'' and in large-scale language models, yet the mechanisms by which gradient-based learning creates the required internal…

Machine Learning · Statistics 2026-05-19 Naman Agarwal , Siddhartha R. Dalal , Vishal Misra

The incredible success of transformers on sequence modeling tasks can be largely attributed to the self-attention mechanism, which allows information to be transferred between different parts of a sequence. Self-attention allows…

Machine Learning · Computer Science 2024-08-14 Eshaan Nichani , Alex Damian , Jason D. Lee

Lateral inhibitory connections have been observed in the cortex of the biological brain, and has been extensively studied in terms of its role in cognitive functions. However, in the vanilla version of backpropagation in deep learning, all…

Computer Vision and Pattern Recognition · Computer Science 2022-08-16 Lei Jiang , Yongqing Liu , Shihai Xiao , Yansong Chua

Large language models frequently exhibit hallucinations: fluent and confident outputs that are factually incorrect or unsupported by the input context. While recent hallucination detection methods have explored various features derived from…

Computation and Language · Computer Science 2026-04-14 Jakub Binkowski , Kamil Adamczewski , Tomasz Kajdanowicz

Language Models (LMs) assign significant attention to the first token, even if it is not semantically important, which is known as attention sink. This phenomenon has been widely adopted in applications such as streaming/long context…

Computation and Language · Computer Science 2025-03-04 Xiangming Gu , Tianyu Pang , Chao Du , Qian Liu , Fengzhuo Zhang , Cunxiao Du , Ye Wang , Min Lin

Large Language Models (LLMs) often allocate disproportionate attention to specific tokens, a phenomenon commonly referred to as the attention sink. While such sinks are generally considered detrimental, prior studies have identified a…

Machine Learning · Computer Science 2026-03-10 Runyu Peng , Ruixiao Li , Mingshu Chen , Yunhua Zhou , Qipeng Guo , Xipeng Qiu

While backpropagation--reverse-mode automatic differentiation--has been extraordinarily successful in deep learning, it requires two passes (forward and backward) through the neural network and the storage of intermediate activations.…

Machine Learning · Computer Science 2025-11-06 Daniel Wang , Evan Markou , Dylan Campbell

Pre-trained Transformers often exhibit over-confidence in source patterns and difficulty in forming new target-domain patterns during fine-tuning. We formalize the mechanism of output saturation leading to gradient suppression through…

Machine Learning · Computer Science 2025-11-04 Wang Zixian

Although transformer-based models have shown exceptional empirical performance, the fundamental principles governing their training dynamics are inadequately characterized beyond configuration-specific studies. Inspired by empirical…

Machine Learning · Computer Science 2025-10-09 Zheng-An Chen , Tao Luo

As the foundational architecture of modern machine learning, Transformers have driven remarkable progress across diverse AI domains. Despite their transformative impact, a persistent challenge across various Transformers is Attention Sink…

Self-attention is the key mechanism of transformers, which are the essential building blocks of modern foundation models. Recent studies have shown that pure self-attention suffers from an increasing degree of rank collapse as depth…

Machine Learning · Computer Science 2024-11-04 Xinyi Wu , Amir Ajorlou , Yifei Wang , Stefanie Jegelka , Ali Jadbabaie

Transformer models have emerged as fundamental tools across various scientific and engineering disciplines, owing to their outstanding performance in diverse applications. Despite this empirical success, the theoretical foundations of…

Machine Learning · Computer Science 2026-04-14 Zhen Qin , Jinxin Zhou , Jiachen Jiang , Zhihui Zhu
‹ Prev 1 2 3 10 Next ›