Variance Sensitivity Induces Attention Entropy Collapse and Instability in Transformers

Jonghyun Hong, Sungyoon Lee. Variance Sensitivity Induces Attention Entropy Collapse and Instability in Transformers. In Christos Christodoulopoulos 0001, Tanmoy Chakraborty 0002, Carolyn Rose, Violet Peng, editors, Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, EMNLP 2025, Suzhou, China, November 4-9, 2025. pages 8360-8378, Association for Computational Linguistics, 2025. [doi]

Abstract

Abstract is missing.