Token Assorted: Mixing Latent and Text Tokens for Improved Language Model Reasoning

@misc{tian2025tokenassortedmixing,
      title={Token Assorted: Mixing Latent and Text Tokens for Improved Language  Model Reasoning}, 
      author={Yuandong Tian and Jiantao Jiao and Yingchen Xu and Hanlin Zhu and DiJia Su and Qinqing Zheng},
      year={2025},
      eprint={2502.03275},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2502.03275}, 
}