@misc{indiciae2e43bcb07d9a,
  title = {ADaPT: Token-Level Decoupling for Efficient Large Reasoning Models},
  author = {Tingyun Li and Zishang Jiang and Jinyi Han and Xinyi Wang and Sihang Jiang and Han Xia and Zhaoqian Dai and Shuguang Ma and Fei Yu and Jiaqing Liang and Yanghua Xiao},
  year = {2026},
  url = {https://arxiv.org/abs/2606.19919},
  note = {Source identifier: 2606.19919}
}