@article{adaptivetesttimecomputeallocationforreas, title = {Adaptive Test-Time Compute Allocation for Reasoning LLMs via Constrained Policy Optimization}, author = {Zhiyuan Zhai and Bingcong Li and Bingnan Xiao and Ming Li and Xin Wang}, year = {2026}, eprint = {2604.14853}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2604.14853}, }