paper-with-me

홈 › Papers

Normalization in Attention Dynamics

2025-10-24 · Nikita Karagodin, Shu Ge, Yury Polyanskiy, Philippe Rigollet arxiv

We study the effect of normalization schemes on token representations in deep transformers. Modeling their evolution as interacting particles on the sphere, we show that normalization acts as a form of speed regulation. This perspective enables a unified analysis of several schemes -- including Post-LN, Pre-LN, Mix-LN, Peri-LN, nGPT -- revealing how they influence clustering dynamics and representation collapse. Our framework clarifies how different schemes shape token representations across layers and provides a principled basis for comparing them, identifying Peri-LN as a particularly effective choice.

📄 PDF Abstract BibTeX arXiv:2510.22026

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Connecting the geometry and dynamics of many-body complex systems with message passing neural operators

2025-02-21 · Nicholas A. Gabriel, Neil F. Johnson, George Em Karniadakis

The relationship between scale transformations and dynamics established by renormalization group techniques is a cornerstone of modern physical theories, from fluid mechanics to elementary particle physics. Integrating r…

Inductive BiasOperator learning

Improving Adversarial Robustness of Attribution via Implicit Regularization

2026-05-28 · Amir Mehrpanah, Matteo Gamba, Hossein Azizpour arxiv

The adversarial robustness of attributions is a fundamental requirement for reliable explainability in deep learning, yet existing approaches typically rely on computationally expensive explicit regularization. In this w…

Adversarial Robustness

The Transformer as a Polar State Estimator

2026-05-10 · Peter Racioppo arxiv

We show that the core components of the Transformer---attention, residual connections, and normalization---arise naturally from a single geometric state estimation problem. Modeling the latent state in polar coordinates …

The Mean-Field Dynamics of Transformers

2025-12-01 · Philippe Rigollet arxiv

We develop a mathematical framework that interprets Transformer attention as an interacting particle system and studies its continuum (mean-field) limits. By idealizing attention on the sphere, we connect Transformer dyn…

Muon Dynamics as a Spectral Wasserstein Flow

2026-04-06 · Gabriel Peyré arxiv

Gradient normalization stabilizes deep-learning optimization, and spectral normalizations are especially natural for matrix-shaped parameter blocks; Muon is the motivating example. We study an idealized deterministic, co…