@misc{indiciaeae5586394893,
  title = {Accelerating Dense LLMs via L0-regularized Mixture-of-Experts},
  author = {Zhenyu Zhang and Jiudong Yang and Zhaowen Tao and Meng Chen},
  year = {2026},
  url = {https://arxiv.org/abs/2609.21672},
  note = {Source identifier: 2609.21672}
}