@article{disaofflineimportancesamplingfordistribu, title = {DISA: Offline Importance Sampling for Distribution-Matching LLM-RL}, author = {Shaobo Wang and Yujie Chen and Yafeng Sun and Wenjie Qiu and Zhihui Xie and Sihang Li and Yucheng Li and Huiqiang Jiang and Xingzhang Ren and Xuming Hu and Dayiheng Liu and Linfeng Zhang}, year = {2026}, eprint = {2605.17295}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2605.17295}, }