@article{pipelineparallelismisallyouneedforoptimi,
  title = {Pipeline Parallelism is All You Need for Optimized Early-Exit Based Self-Speculative Decoding},
  author = {Ruanjun Li and Ziheng Liu and Yuanming Shi and Jiawei Shao and Chi Zhang and Xuelong Li},
  year = {2025},
  eprint = {2509.19368},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2509.19368},
}