@misc{indiciae94e5bd9a3ca3,
  title = {Erase to Improve: Erasable Reinforcement Learning for Search-Augmented LLMs},
  author = {Ziliang Wang and Kang An and Xuhui Zheng and Faqiang Qian and Weikun Zhang and Cijun Ouyang and Jialu Cai and Yuhang Wang and Yichao Wu},
  year = {2026},
  url = {https://arxiv.org/abs/2510.00861},
  note = {Source identifier: 2510.00861}
}