@inproceedings{zheng-etal-2026-gradients,
    title = "Gradients Know What Outcomes Don{'}t: Unlocking Reinforcement Learning for {LLM} Reasoning with Gradient-Aligned Rewards",
    author = "Zheng, Leqi  and
      Su, Jinbo  and
      Niu, Fang  and
      Wang, Chaokun  and
      Wang, Weiping  and
      Zhang, Jiajun  and
      Yan, Shannan  and
      Wu, Jie  and
      Kang, Zhaolu  and
      Fu, Rong  and
      Zhang, Hang",
    editor = "Lee, Hung-Yi  and
      Dev, Sunipa  and
      Li, Jessy  and
      Elliott, Desmond",
    booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
    month = oct,
    year = "2026",
    address = "Budapest, Hungary",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2026.emnlp-main.594/",
    pages = "13686--13706",
    ISBN = "979-8-89176-459-0"
}