Abstract
What is happiness for reinforcement learning agents? We seek a formal definition satisfying a list of desiderata. Our proposed definition of happiness is the temporal difference error, i.e. the difference between the value of the obtained reward and observation and the agent’s expectation of this value. This definition satisfies most of our desiderata and is compatible with empirical research on humans. We state several implications and discuss examples.
Research supported by the People for the Ethical Treatment of Reinforcement Learners http://petrl.org. See the extended technical report for omitted proofs and details about the data analysis [4].
Both authors contributed equally.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Similar content being viewed by others
References
Bostrom, N.: Superintelligence: Paths, Dangers. Oxford University Press, Strategies (2014)
Brickman, P., Campbell, D.T.: Hedonic relativism and planning the good society. Adaptation-Level Theory, pp. 287–305 (1971)
Brickman, P., Coates, D., Janoff-Bulman, R.: Lottery winners and accident victims: Is happiness relative? Journal of Personality and Social Psychology 36, 917 (1978)
Daswani, M., Leike, J.: A definition of happiness for reinforcement learning agents. Technical report, Australian National University (2015). http://arxiv.org/abs/1505.04497
Diener, E., Lucas, R.E., Scollon, C.N.: Beyond the hedonic treadmill: Revising the adaptation theory of well-being. American Psychologist 61, 305 (2006)
Jacobs, E., Broekens, J., Jonker, C.: Joy, distress, hope, and fear in reinforcement learning. In: Conference on Autonomous Agents and Multiagent Systems, pp. 1615–1616 (2014)
Niv, Y.: Reinforcement learning in the brain. Journal of Mathematical Psychology 53, 139–154 (2009)
Rutledge, R.B., Skandali, N., Dayan, P., Dolan, R.J.: A computational and neural model of momentary subjective well-being. In: Proceedings of the National Academy of Sciences (2014)
Schmidhuber, J.: Formal theory of creativity, fun, and intrinsic motivation (1990–2010). IEEE Transactions on Autonomous Mental Development. 2, 230–247 (2010)
Sutton, R., Barto, A.: Time-derivative models of Pavlovian reinforcement. In: Learning and Computational Neuroscience: Foundations of Adaptive Networks, pp. 497–537. MIT Press (1990)
Sutton, R.S., Barto, A.G.: Reinforcement Learning: An Introduction. MIT Press, Cambridge (1998)
Tomasik, B.: Do artificial reinforcement-learning agents matter morally? Technical report, Foundational Research Institute (2014). http://arxiv.org/abs/1410.8233
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2015 Springer International Publishing Switzerland
About this paper
Cite this paper
Daswani, M., Leike, J. (2015). A Definition of Happiness for Reinforcement Learning Agents. In: Bieger, J., Goertzel, B., Potapov, A. (eds) Artificial General Intelligence. AGI 2015. Lecture Notes in Computer Science(), vol 9205. Springer, Cham. https://doi.org/10.1007/978-3-319-21365-1_24
Download citation
DOI: https://doi.org/10.1007/978-3-319-21365-1_24
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-21364-4
Online ISBN: 978-3-319-21365-1
eBook Packages: Computer ScienceComputer Science (R0)