Abstract
Ability to use language is an essential requirement for human-level intelligence. For artificial general intelligence, the ability to learn and to create language is even more important [1]. Most previous models of learning and emergence of language took successful communication itself as the task target. However, language, or communication in general, should have evolved to improve certain fitness of the population of agents. Here we consider whether and how a population of reinforcement learning agents can learn to send signals and to respond to signals for the sake of maximizing their own rewards. We take a communication game tested in human subjects [2, 3, 6], in which the aim of the game is for two players to meet together without knowing exact location of the other. In our decentralized reinforcement learning framework with communicative and physical actions [4], we tested how the number N of usable symbols affects whether the meeting task is successfully achieved and what kind of signaling and responding are learned. Even though \(N=2\) symbols are theoretically sufficient, the success rate was only 1 to 2%. With \(N=3\) symbols, success rate was more than 60% and three different signaling strategies were observed. The results indicate the importance of redundancy in signaling degrees of freedom and that a variety of signaling conventions can emerge in populations of simple independent reinforcement learning agents.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Doya, K., Taniguchi, T.: Toward evolutionary and developmental intelligence. Curr. Opin. Behav. Sci. 29, 91–96 (2019)
Galantucci, B.: An experimental study of the emergence of human communication systems. Cogn. Sci. 29(5), 737–67 (2005)
Galantucci, B., Steels, L.: The emergence of embodied communication in artificial agents and humans. In: Wachsmuth, I., Lenzen, M., Knoblich, G. (eds.) Embodied Communication in Humans and Machines, chap. 11, pp. 229–256. Oxford University Press, Oxford (2008)
Huang, Q., Uchibe, E., Doya, K.: Emergence of communication among reinforcement learning agents under coordination environment. In: The Sixth Joint IEEE International Conference Developmental Learning and Epigenetic Robotics, pp. 57–58 (2016)
Klein, M., Kamp, H., Palm, G., Doya, K.: A computational neural model of goal-directed utterance selection. Neural Netw. 23(5), 592–606 (2010)
Konno, T., Morita, J., Hashimoto, T.: Symbol communication systems integrate implicit information in coordination tasks. In: Dietterich, T., Becker, S., Ghahramani, Z. (eds.) Advances in Cognitive Neurodynamics (III), pp. 453–459. Springer, Dordrecht (2013). https://doi.org/10.1007/978-94-007-4792-0_61
Mordatch, I., Abbeel, P.: Emergence of grounded compositional language in multi-agent populations. In: Thirty-Second AAAI Conference on Artificial Intelligence (2018)
Sato, T., Uchibe, E., Doya, K.: Learning how, what, and whether to communicate: emergence of protocommunication in reinforcement learning agents. Artif. Life Robot. 12, 70–74 (2008)
Sutton, R.S., Barto, A.G.: Reinforcement Learning: An Introduction. MIT Press, Cambridge (2018)
Acknowledgements
This work was supported by Ministry of Education, Culture, Sports, Science, and Technology KAKENHI Grants 23120007 and 16H06563, and research support of Okinawa Institute of Science and Technology Graduate University to KD.
Author information
Authors and Affiliations
Corresponding authors
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2019 Springer Nature Switzerland AG
About this paper
Cite this paper
Huang, Q., Kenji, D. (2019). An Experimental Study of Emergence of Communication of Reinforcement Learning Agents. In: Hammer, P., Agrawal, P., Goertzel, B., Iklé, M. (eds) Artificial General Intelligence. AGI 2019. Lecture Notes in Computer Science(), vol 11654. Springer, Cham. https://doi.org/10.1007/978-3-030-27005-6_9
Download citation
DOI: https://doi.org/10.1007/978-3-030-27005-6_9
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-27004-9
Online ISBN: 978-3-030-27005-6
eBook Packages: Computer ScienceComputer Science (R0)