Conference item icon

Conference item

Trial without error: Towards safe reinforcement learning via human intervention

Alternative title:
Extended abstract
Abstract:

During training, model-free reinforcement learning (RL) systems can explore actions that lead to harmful or costly consequences. Having a human “in the loop” and ready to intervene at all times can prevent these mistakes, but is prohibitively expensive for current algorithms. We explore how human oversight can be combined with a supervised learning system to prevent catastrophic events during training. We demonstrate this scheme on Atari games, with a Deep RL agent being overseen by a human f...

Expand abstract
Publication status:
Published
Peer review status:
Peer reviewed

Actions


Access Document


Files:

Authors


More by this author
Institution:
University of Oxford
Role:
Author
More by this author
Institution:
University of Oxford
Role:
Author
More by this author
Institution:
University of Oxford
Division:
HUMS
Department:
Philosophy Faculty
Oxford college:
St John's College
Role:
Author
Publisher:
ACM Digital Library Publisher's website
Journal:
17th International Conference on Autonomous Agents and MultiAgent Systems Journal website
Host title:
Proceedings of the 17th International Conference on Autonomous Agents and MultiAgent Systems
Publication date:
2018-07-15
Acceptance date:
2018-04-15
ISSN:
2523-5699
Source identifiers:
966914
Pubs id:
pubs:966914
UUID:
uuid:ba0fc61e-1a94-4738-929b-cdfd92fdf49e
Local pid:
pubs:966914
Deposit date:
2019-01-30

Terms of use


Views and Downloads






If you are the owner of this record, you can report an update to it here: Report update to this record

TO TOP