@misc{indiciae3e19d7ee7b09,
  title = {ImproveAnyTask: An Autonomous Post-Training Harness for Iterative Model Self-Improvement},
  author = {Xingbo Yao and Xiaoman Wang and Zhengwu Lei and Tinghui Luo and YiLin Zhang and Yuefeng Wu and Yijie Xu and Tianfu Wang and Qingyuan Zhan and Ye Guo and Daoxin Zhang and Zhe Xu and Jian Liu and Hui Xiong},
  year = {2026},
  url = {https://arxiv.org/abs/2610.06347},
  note = {Source identifier: 2610.06347}
}