@article{theeasythehardandthelearnableconfidencea,
  title = {The Easy, the Hard, and the Learnable: Confidence and Difficulty-Adaptive Policy Optimization for LLM Reasoning},
  author = {Zhanke Zhou and Xiangyu Lu and Chentao Cao and Brando Miranda and Tongliang Liu and Bo Han and Sanmi Koyejo},
  year = {2026},
  eprint = {2606.07950},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2606.07950},
}