What drives attention sinks? A study of massive activations and rotational positional encoding in large vision-language models

Open Webpage

Xiaofeng Zhang, Yuanchao Zhu, Chaochen Gu, Jiawei Cao, Hao Cheng, Kaijie Wu

Published: 2026, Last Modified: 28 Feb 2026Inf. Process. Manag. 2026EveryoneRevisionsBibTeXCC BY-SA 4.0