@article{advantagecollapseingrouprelativepolicyop, title = {Advantage Collapse in Group Relative Policy Optimization: Diagnosis and Mitigation}, author = {Xixiang He and Qiyao Sun and Ao Cheng and Xingming Li and Xuanyu Ji and Hailun Lu and Runke Huang and Qingyong Hu}, year = {2026}, eprint = {2605.21125}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2605.21125}, }