@misc{indiciae093b8ab2be54,
  title = {SafeExplorer: An Unbiased Policy Gradient for Reinforcement Learning with Recovery Interventions},
  author = {Elham Daneshmand and Majid Khadiv and Glen Berseth and Hsiu-Chin Lin},
  year = {2026},
  url = {https://arxiv.org/abs/2607.08925},
  note = {Source identifier: 2607.08925}
}