Poster "intra-modality modeling" Papers

Accelerating Multimodal Large Language Models via Dynamic Visual-Token Exit and the Empirical Findings

Qiong Wu, Wenhao Lin, Yiyi Zhou et al.