paper-with-me

Papers

Is Anisotropy Inherent to Transformers?

2023-06-13 · Nathan Godey, Éric de la Clergerie, Benoît Sagot

The representation degeneration problem is a phenomenon that is widely observed among self-supervised learning methods based on Transformers. In NLP, it takes the form of anisotropy, a singular property of hidden representations which makes them unexpectedly close to each other in terms of angular distance (cosine-similarity). Some recent works tend to show that anisotropy is a consequence of optimizing the cross-entropy loss on long-tailed distributions of tokens. We show in this paper that anisotropy can also be observed empirically in language models with specific objectives that should not suffer directly from the same consequences. We also show that the anisotropy problem extends to Transformers trained on other modalities. Our observations tend to demonstrate that anisotropy might actually be inherent to Transformers-based models.

📄 PDF Abstract BibTeX arXiv:2306.07656

Code (0)

등록된 구현이 없습니다.

Tasks

Self-Supervised Learning

Similar Papers 제목 키워드 기반

Anisotropy Is Inherent to Self-Attention in Transformers

2024-01-22 · Nathan Godey, Éric de la Clergerie, Benoît Sagot

The representation degeneration problem is a phenomenon that is widely observed among self-supervised learning methods based on Transformers. In NLP, it takes the form of anisotropy, a singular property of hidden represe…

Self-Supervised Learning

Revisiting Anisotropy in Language Transformers: The Geometry of Learning Dynamics

2026-04-09 · Raphael Bernas, Fanny Jourdan, Antonin Poché, Céline Hudelot arxiv

Since their introduction, Transformer architectures have dominated Natural Language Processing (NLP). However, recent research has highlighted an inherent anisotropy phenomenon in these models, presenting a significant c…

Transformer Geometry Observatory TGO-I: Spectral Geometry Observatory

2026-06-17 · Kaustubh Kapil, Kishor P. Upla arxiv

Despite the widespread adoption of Vision Transformers (ViTs) and their success across numerous computer vision applications, the fundamental understanding of their dimensional and representational geometry remains relat…

Introducing Anisotropic Minkowski Functionals and Quantitative Anisotropy Measures for Local Structure Analysis in Biomedical Imaging

2020-04-02 · Axel Wismueller, Titas De, Eva Lochmueller, Felix Eckstein 외

The ability of Minkowski Functionals to characterize local structure in different biological tissue types has been demonstrated in a variety of medical image processing tasks. We introduce anisotropic Minkowski Functiona…

Pressure and temperature dependence of fluorescence anisotropy of Green Fluorescent Protein

2022-02-15 · Harpreet Kaur, Khanh Nguyen, Pradeep Kumar

We have studied the effect of high hydrostatic pressure and temperature on the steady state fluorescence anisotropy of Green Fluorescent Protein (GFP). We find that the fluorescence anisotropy of GFP at a constant temper…