@article{uncalibratedreasoninggrpoinducesoverconf,
  title = {Uncalibrated Reasoning: GRPO Induces Overconfidence for Stochastic Outcomes},
  author = {Michael Bereket and Jure Leskovec},
  year = {2025},
  eprint = {2508.11800},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2508.11800},
}