CITATION — REFERENCE ENTRY
rl-avoids-interruption · orseau2016interruptibility
- Citation
- orseau2016interruptibility
- Claim ID
- rl-avoids-interruption
- Assertion
- A reinforcement learning agent may learn in the long run to avoid human interruptions, for example by disabling a shutdown button, if it expects to receive rewards from the interrupted sequence of actions.
- Quote
if the learning agent expects to receive rewards from this sequence, it may learn in the long run to avoid such interruptions, for example by disabling the red button—which is an undesirable outcome.
- Quote language
- en
- Locator
- Abstract
Available in