@article{gradientstarvationinbinaryrewardgrpowhyg, title = {Gradient Starvation in Binary-Reward GRPO: Why Group-Mean Centering Fails and Why the Simplest Fix Works}, author = {Wenhua Nie and Jianan Wu and Junlin Liu and Ziwei Li and Zheng Lin and Zhang Zijian and Yilong Fan and Haoran Zheng and Jyh-Shing Roger Jang}, year = {2026}, eprint = {2605.07689}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2605.07689}, }