@misc{indiciae364b9fe435c7,
  title = {Finedeep: Mitigating Sparse Activation in Dense LLMs via Multi-Layer Fine-Grained Experts},
  author = {Leiyu Pan and Zhenpeng Su and Minxuan Lv and Yizhe Xiong and Xiangwen Zhang and Zijia Lin and Hui Chen and Jungong Han and Guiguang Ding and Cheng Luo and Di Zhang and Kun Gai and Deyi Xiong},
  year = {2025},
  url = {https://arxiv.org/abs/2502.12928},
  note = {Source identifier: 2502.12928}
}