@article{thesharpnessdisparityprinciplein,
  title = {The Sharpness Disparity Principle in Transformers for Accelerating Language Model Pre-Training},
  author = {Jinbo Wang and Mingze Wang and Zhanpeng Zhou and Junchi Yan and Weinan E and Lei Wu},
  year = {2025},
  eprint = {2502.19002},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2502.19002v2},
}