@article{evocomplearningvisualtokencompressionfor,
  title = {EvoComp: Learning Visual Token Compression for Multimodal Large Language Models via Semantic-Guided Evolutionary Labeling},
  author = {Jiafei Song and Fengwei Zhou and Jin Qu and Wenjin Jason Li and Tong Wu and Gengjian Xue and Zhikang Zhao and Daomin Wei and Yichao Lu and Bailin Na},
  year = {2026},
  eprint = {2604.17087},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2604.17087},
}