Artifacts and Attention Sinks: Structured Approximations for Efficient Vision Transformers
Vision transformers have emerged as a powerful tool across a wide range of applications, yet their inner workings remain only partially understood. In this work, we examine the phenomenon of massive tokens - tokens with exceptionally high activation norms that act as attention sinks - and artifact tokens that emerge as a byproduct during inference. Our analysis reveals that these tokens mutually suppress one another through the attention mechanism, playing a critical role in regulating information flow within the network. Leveraging these insights, we introduce Fast Nyström Attention (FNA), a training-free method that approximates self-attention in linear time and space by exploiting the structured patterns formed by massive and artifact tokens. Additionally, we propose a masking strategy to mitigate noise from these tokens, yielding modest performance gains at virtually no cost. We evaluate our approach on popular pretrained vision backbones and demonstrate competitive performance on retrieval, classification, segmentation, and visual question answering (VQA), all while reducing computational overhead.
Code (0)
등록된 구현이 없습니다.
Tasks
Visual Question AnsweringSimilar Papers 제목 키워드 기반
When Sinks Help or Hurt: Unified Framework for Attention Sink in Large Vision-Language Models
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 largel…
On the Existence and Behavior of Secondary Attention Sinks
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 t…
A Unifying View of Attention Sinks: Two Algorithms, Two Solutions
When attention concentrates on a single token, a sink, what is the model actually computing? Attention sinks are ubiquitous in softmax transformers, yet this shared visual signature can hide fundamentally different algor…
Sink vs. diagonal patterns as mechanisms for attention switch and oversmoothing prevention
This paper studies the role of sinks and diagonal patterns as attention switch and anti-oversmoothing mechanisms. We analyze geometric conditions under which sinks can be represented, showing a necessary alignment betwee…
Seeing Clearly by Layer Two: Enhancing Attention Heads to Alleviate Hallucination in LVLMs
The hallucination problem in multimodal large language models (MLLMs) remains a common issue. Although image tokens occupy a majority of the input sequence of MLLMs, there is limited research to explore the relationship …
Hallucination