[2501.18322v1] A Unified Perspective on the Dynamics of Deep Transformers

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs. arXiv Operational Status Get status notifications via email or slack

[2501.18322v1] A Unified Perspective on the Dynamics of Deep Transformers Skip to main content arXiv is now an independent nonprofit! Learn more × Search arXiv Press Enter to search · Advanced search --> Computer Science > Machine Learning arXiv:2501.18322v1 (cs) [Submitted on 30 Jan 2025 (this version), latest version 17 Jun 2026 ( v2 )] Title: A Unified Perspective on the Dynamics of Deep Transformers Authors: Valérie Castin , Pierre Ablin , José Antonio Carrillo , Gabriel Peyré View a PDF of the paper titled A Unified Perspective on the Dynamics of Deep Transformers, by Val\'er

related reading

• [2512.19428] Attention Is Not What You Needarxiv.org
• [1807.03819] Universal Transformersarxiv.org
• [2508.12773] Online Ensemble Transformer for Accurate Cloud Workload Forecasting in Predictive Auto-Scalingarxiv.org
• Scalable Diffusion Models with Transformersarxiv.org
• [2305.18654] Faith and Fate: Limits of Transformers on Compositionalityarxiv.org
• [2201.06618] VAQF: Fully Automatic Software-Hardware Co-Design Framework for Low-Bit Vision Transformerarxiv.org
• [2401.02721] A Cost-Efficient FPGA Implementation of Tiny Transformer Model using Neural ODEarxiv.org
• [2402.12875] Chain of Thought Empowers Transformers to Solve Inherently Serial Problemsarxiv.org
• [2406.11717] Refusal in Language Models Is Mediated by a Single Directionarxiv.org
• [2506.06194] Transformative or Conservative? Conservation laws for ResNets and Transformersarxiv.org
• [2207.00032] DeepSpeed Inference: Enabling Efficient Inference of Transformer Models at Unprecedented Scalearxiv.org
• [2504.18208v1] Ultra-fast feature learning for the training of two-layer neural networks in the two-timescale regimearxiv.org