@misc{indiciae706aafe8ceaf,
  title = {Semantic Parallelism: Redefining Efficient MoE Inference via Model-Data Co-Scheduling},
  author = {Yan Li and Zhenyu Zhang and Zhengang Wang and Pengfei Chen and Pengfei Zheng},
  year = {2026},
  url = {https://arxiv.org/abs/2503.04398},
  note = {Source identifier: 2503.04398}
}