Looping Back to Move Forward: Recursive Transformers for Efficient and Flexible Large Multimodal Models

Large Multimodal Models (LMMs) have achieved remarkable success in vision-language tasks, yet their vast parameter counts are often underutilized during both training and inference. In this work, we embrace the idea of looping back to move forward: reusing model parameters through recursive refineme...

Descripción completa

Detalles Bibliográficos

Autores principales:	Xu, Ruihan, Gao, Yuting, Wang, Lan, Li, Jianing, Chen, Weihao, Guo, Qingpei, Yang, Ming, Zhang, Shiliang
Formato:	Artículo
Publicado:	2026
Materias:	Machine Learning Artificial Intelligence
Acceso en línea:	https://arxiv.org/abs/2602.09080 http://arxiv.org/abs/2602.09080
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!