@misc{indiciaedf025e60d59b,
  title = {Off-Policy Reinforcement Learning with Anytime Safety Guarantees via Robust Safe Gradient Flow},
  author = {Pol Mestres and Arnau Marzabal and Jorge Cortés},
  year = {2025},
  url = {https://arxiv.org/abs/2510.01492},
  note = {Source identifier: 2510.01492}
}