@misc{indiciae9a4d520de0f0,
  title = {Toward Skill-Native LLMs: Skill Entropy for Benchmarking and Training Long-Horizon Reasoning},
  author = {Yinghui He and Ling Yang and Jiarui Liu and Yongjin Yang and Lechen Zhang and Yingcheng Wu and Zhenfei Yin and Mengdi Wang and Sanjeev Arora},
  year = {2026},
  url = {https://arxiv.org/abs/2608.05139},
  note = {Source identifier: 2608.05139}
}