paper-with-me

홈 › Papers

What are you sinking? A geometric approach on attention sink

2025-08-04 · Valeria Ruscio, Umberto Nanni, Fabrizio Silvestri arxiv

Attention sink (AS) is a consistent pattern in transformer attention maps where certain tokens (often special tokens or positional anchors) disproportionately attract attention from other tokens. We show that in transformers, AS is not an architectural artifact, but it is the manifestation of a fundamental geometric principle: the establishment of reference frames that anchor representational spaces. We analyze several architectures and identify three distinct reference frame types, centralized, distributed, and bidirectional, that correlate with the attention sink phenomenon. We show that they emerge during the earliest stages of training as optimal solutions to the problem of establishing stable coordinate systems in high-dimensional spaces. We show the influence of architecture components, particularly position encoding implementations, on the specific type of reference frame. This perspective transforms our understanding of transformer attention mechanisms and provides insights for both architecture design and the relationship with AS.

📄 PDF Abstract BibTeX arXiv:2508.02546

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Spectral Filters, Dark Signals, and Attention Sinks

2024-02-14 · Nicola Cancedda

Projecting intermediate representations onto the vocabulary is an increasingly popular interpretation tool for transformer-based LLMs, also known as the logit lens. We propose a quantitative extension to this approach an…

Attention Sinks in Diffusion Language Models

2025-10-17 · Maximo Eduardo Rulli, Simone Petruzzi, Edoardo Michielon, Fabrizio Silvestri 외 arxiv

Masked Diffusion Language Models (DLMs) have recently emerged as a promising alternative to traditional Autoregressive Models (ARMs). DLMs employ transformer encoders with bidirectional attention, enabling parallel token…

Sink vs. diagonal patterns as mechanisms for attention switch and oversmoothing prevention

2026-05-08 · Peter Súkeník, Cristina López Amado, Christoph H. Lampert, Marco Mondelli arxiv

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…

Computer vision-based estimation of invertebrate biomass

2026-03-06 · Mikko Impiö, Philipp M. Rehsen, Jarrett Blair, Cecilie Mielec 외 arxiv

The ability to estimate invertebrate biomass using only images could help scaling up quantitative biodiversity monitoring efforts. Computer vision-based methods have the potential to omit the manual, time-consuming, and …

Data Augmentation

SimuScene: Simulation-Ready Compositional 3D Scene Reconstruction from a Single Image

2026-06-02 · Inhee Lee, Sangwon Baik, Sungjoo Kim, Hyeonwoo Kim 외 arxiv

Reconstructing interactive, simulation-ready 3D scenes from a single image is a critical bottleneck for robotic manipulation. While recent single-image lifters recover plausible per-object shapes, composing them yields s…

3D Reconstruction