English
Related papers

Related papers: Mitigating Attention Sinks and Massive Activations…

200 papers

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

Audio-Visual Speech Recognition (AVSR) achieves robust speech recognition in noisy environments by combining auditory and visual information. However, recent Large Language Model (LLM) based AVSR systems incur high computational costs due…

Computer Vision and Pattern Recognition · Computer Science 2025-06-06 Jeong Hun Yeo , Hyeongseop Rha , Se Jin Park , Yong Man Ro

Large multimodal models (LMMs) "see" images by leveraging the attention mechanism between text and visual tokens in the transformer decoder. Ideally, these models should focus on key visual information relevant to the text token. However,…

Computer Vision and Pattern Recognition · Computer Science 2025-03-06 Seil Kang , Jinyeong Kim , Junhyeok Kim , Seong Jae Hwang

Multimodal large language models (MLLMs) have recently become a focal point of research due to their formidable multimodal understanding capabilities. For example, in the audio and speech domains, an LLM can be equipped with (automatic)…

Computer Vision and Pattern Recognition · Computer Science 2025-03-10 Umberto Cappellazzo , Minsu Kim , Honglie Chen , Pingchuan Ma , Stavros Petridis , Daniele Falavigna , Alessio Brutti , Maja Pantic

Practitioners have consistently observed three puzzling phenomena in transformer-based large language models (LLMs): attention sinks, value-state drains, and residual-state peaks, collectively referred to as extreme-token phenomena. These…

Machine Learning · Computer Science 2024-11-08 Tianyu Guo , Druv Pai , Yu Bai , Jiantao Jiao , Michael I. Jordan , Song Mei

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

Attention sinks are tokens, often the beginning-of-sequence (BOS) token, that receive disproportionately high attention despite limited semantic relevance. In this work, we identify a class of attention sinks, which we term secondary sinks,…

Machine Learning · Computer Science 2026-03-17 Jeffrey T. H. Wong , Cheng Zhang , Louis Mahon , Wayne Luk , Anton Isopoussu , Yiren Zhao

The goal of this paper is to strengthen the reasoning of Omnimodal Large Language Models (Omni-LLMs) at inference time, without additional training. These models jointly process video, audio, and text, and given the large number of tokens…

Computer Vision and Pattern Recognition · Computer Science 2026-05-08 Suho Yoo , Youngjoon Jang , Joon Son Chung

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…

Large Vision Language Models (LVLMs) have recently emerged as powerful architectures capable of understanding and reasoning over both visual and textual information. These models typically rely on two key components: a Vision Transformer…

Computer Vision and Pattern Recognition · Computer Science 2025-10-10 Jiayun Luo , Wan-Cyuan Fan , Lyuyang Wang , Xiangteng He , Tanzila Rahman , Purang Abolmaesumi , Leonid Sigal

We investigate the origins of massive activations in large language models (LLMs) and identify a specific layer named the \textbf{Massive Emergence Layer (ME Layer)}, that is consistently observed across model families, where massive…

Computation and Language · Computer Science 2026-05-14 Zeru Shi , Zhenting Wang , Fan Yang , Qifan Wang , Ruixiang Tang

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

Multimodal large language models (MLLMs) demand considerable computations for inference due to the extensive parameters and the additional input tokens needed for visual information representation. Herein, we introduce Visual Tokens…

Computer Vision and Pattern Recognition · Computer Science 2025-01-28 Zhihang Lin , Mingbao Lin , Luxi Lin , Rongrong Ji

Large Language Models (LLMs) often assign disproportionate attention to the first token, a phenomenon known as the attention sink. Several recent approaches aim to address this issue, including Sink Attention in GPT-OSS and Gated Attention…

Computation and Language · Computer Science 2026-05-28 Zizhuo Fu , Wenxuan Zeng , Runsheng Wang , Meng Li

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

Large language models (LLMs) have shown promising efficacy across various tasks, becoming powerful tools in numerous aspects of human life. However, Transformer-based LLMs suffer a performance degradation when modeling long-term contexts…

Computation and Language · Computer Science 2026-03-23 Weiyao Luo , Suncong Zheng , Heming Xia , Weikang Wang , Yan Lei , Tianyu Liu , Shuang Chen , Zhifang Sui

Large language models (LLMs) suffer from hallucination and context forgetting. Prior studies suggest that attention drift is a primary cause of these problems, where LLMs' focus shifts towards newly generated tokens and away from the…

Computer Vision and Pattern Recognition · Computer Science 2026-05-19 Xu Liu , Guikun Chen , Wenguan Wang

Large Language Models (LLMs) are known to contain significant redundancy, yet a systematic explanation for why certain components, particularly in higher layers, are more redundant has remained elusive. In this work, we identify the BOS…

Computation and Language · Computer Science 2026-01-13 Jaewon Sok , Jewon Yeom , Seonghyeon Park , Jeongjae Park , Taesup Kim

In real-world applications, automatic speech recognition (ASR) systems must handle overlapping speech from multiple speakers and recognize rare words like technical terms. Traditional methods address multi-talker ASR and contextual biasing…

Audio and Speech Processing · Electrical Eng. & Systems 2025-06-17 Jiajun He , Naoki Sawada , Koichi Miyazaki , Tomoki Toda

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
‹ Prev 1 2 3 10 Next ›