AIGO
ASVR. Autoregressive Semantic Visual Reconstruction Helps VLMs Understand Better
UniReason. Python
MoIIE. MoIIE: Mixture of Intra- and Inter-Modality Experts for Large Vision Language Models