paper-with-me

홈 › Papers

Addressing Token Uniformity in Transformers via Singular Value Transformation

2022-08-24 · Hanqi Yan, Lin Gui, Wenjie Li, Yulan He

Token uniformity is commonly observed in transformer-based models, in which different tokens share a large proportion of similar information after going through stacked multiple self-attention layers in a transformer. In this paper, we propose to use the distribution of singular values of outputs of each transformer layer to characterise the phenomenon of token uniformity and empirically illustrate that a less skewed singular value distribution can alleviate the `token uniformity' problem. Base on our observations, we define several desirable properties of singular value distributions and propose a novel transformation function for updating the singular values. We show that apart from alleviating token uniformity, the transformation function should preserve the local neighbourhood structure in the original embedding space. Our proposed singular value transformation function is applied to a range of transformer-based language models such as BERT, ALBERT, RoBERTa and DistilBERT, and improved performance is observed in semantic textual similarity evaluation and a range of GLUE tasks. Our source code is available at https://github.com/hanqi-qi/tokenUni.git.

📄 PDF Abstract BibTeX arXiv:2208.11790

Code (1)

hanqi-qi/tokenuni 공식 구현 jax

Tasks

Semantic Textual Similarity

Methods 이 논문이 사용한 방법론

Attention 설명 없음
Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Layer Normalization Unlike batch normalization, Layer Normalization directly estimates the normalization statistics from the summed inputs…
Weight Decay 설명 없음
Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…
WordPiece 설명 없음
Attention Dropout Attention Dropout is a type of dropout used in attention-based architectures, where elements are randomly dropped out of the…
LAMB LAMB is a a layerwise adaptive large batch optimization technique. It provides a strategy for adapting the learning rate in large batch settings. LAMB uses…

Similar Papers 제목 키워드 기반

SAM-DCE: Addressing Token Uniformity and Semantic Over-Smoothing in Medical Segmentation

2025-09-21 · Yingzhen Hu, Yiheng Zhong, Ruobing Li, Yingxue Su 외 arxiv

The Segment Anything Model (SAM) demonstrates impressive zero-shot segmentation ability on natural images but encounters difficulties in medical imaging due to domain shifts, anatomical variability, and its reliance on u…

Demystifying Singular Defects in Large Language Models

2025-02-10 · Haoqi Wang, Tong Zhang, Mathieu Salzmann

Large transformer models are known to produce high-norm tokens. In vision transformers (ViTs), such tokens have been mathematically modeled through the singular vectors of the linear approximations of layers. However, in…

Quantization

Dissecting Query-Key Interaction in Vision Transformers

2024-04-04 · Xu Pan, Aaron Philip, Ziqian Xie, Odelia Schwartz

Self-attention in vision transformers is often thought to perform perceptual grouping where tokens attend to other tokens with similar embeddings, which could correspond to semantically similar features of an object. How…

EvoMoE: Expert Evolution in Mixture of Experts for Multimodal Large Language Models

2025-05-28 · Linglin Jing, Yuting Gao, Zhigang Wang, Wang Lan 외

Recent advancements have shown that the Mixture of Experts (MoE) approach significantly enhances the capacity of large language models (LLMs) and improves performance on downstream tasks. Building on these promising resu…

Mixture-of-ExpertsMMETextVQA

Learning Advanced Self-Attention for Linear Transformers in the Singular Value Domain

2025-05-13 · Hyowon Wi, Jeongwhan Choi, Noseong Park

Transformers have demonstrated remarkable performance across diverse domains. The key component of Transformers is self-attention, which learns the relationship between any two tokens in the input sequence. Recent studie…

Time Series Classification