@misc{indiciaea006f1064bdd,
  title = {Same Beliefs, Different Policies: Learning Reusable Representations for Early Stopping in LLM Reasoning},
  author = {Jingwei Ni and Dmitry Abulkhanov and Vladislav Smirnov and Markus Leippold and Elliott Ash and Timothy Baldwin and Artem Shelmanov and Mrinmaya Sachan},
  year = {2026},
  url = {https://arxiv.org/abs/2610.30450},
  note = {Source identifier: 2610.30450}
}