@article{stabilizingpolicyoptimizationvialogitsco, title = {Stabilizing Policy Optimization via Logits Convexity}, author = {Hongzhan Chen and Tao Yang and Yuhua Zhu and Shiping Gao and Xiaojun Quan and Ting Yao}, year = {2026}, eprint = {2603.00963}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2603.00963}, }