None defined yet.
AVTR: Adaptive Multimodal Reasoning with Visual Imagination
Unbundling Multimodal Vision Encoding