1 paper
Ye Xiao, Shesheng Song, Yi Ren +5
Current Multimodal Large Language Models (MLLMs) have advanced rapidly and follow two main paradigms: one involves concatenating visual and textual tokens for input into Large Lang…