Adaptive Agents and Multi-Agent Systems · 2017 · Conference paper

Trial without Error: Towards Safe Reinforcement Learning via Human Intervention

William S. Saunders, Girish Sastry, Andreas Stuhlmüller, Owain Evans

University of Oxford · Stanford University

Date
2018-07-09
Citations
280
arXiv
1707.05173
Cite
@inproceedings{saunders2017trial,
  title = {Trial without Error: Towards Safe Reinforcement Learning via Human Intervention},
  author = {William S. Saunders and Girish Sastry and Andreas Stuhlmüller and Owain Evans},
  year = {2017},
  booktitle = {Adaptive Agents and Multi-Agent Systems},
  eprint = {1707.05173},
  archivePrefix = {arXiv},
  doi = {10.65109/mzmg4429},
  url = {https://arxiv.org/abs/1707.05173},
}