@misc{indiciae79eefa3f07f6,
  title = {Scaling Model-Generated Distillation Data Can Make Latent Teacher Traits More Recoverable},
  author = {Zhichen Dong and Zhixuan Liu and Yuyu Fan and Xiangtian Li and Shuyang Zhang and Chao Yang},
  year = {2026},
  url = {https://arxiv.org/abs/2608.26958},
  note = {Source identifier: 2608.26958}
}