1 paper
Zeyuan Yang, Hao-Wei Chen, Xueyang Yu +5
Recent years have seen remarkable progress in unified vision-language models handling both multimodal understanding and generation within a single architecture. While autoregressiv…