Tabula rasa agents display emergent in-group behavior
Proceedings of the National Academy of Sciences · 24.06.2025 · 122 (25)
DOI: 10.1073/pnas.2319947121 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the National Academy of Sciences, online erschienen am 16.06.2025. Verfasst von Raphael Köster und 3 weiteren Autor:innen. Beteiligte Einrichtung: Google DeepMind, London EC4A 3TW, United Kingdom. Hauptthemen laut MEDLINE: Social Behavior, Group Processes. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 71 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Theories on group-bias often posit an internal preparedness to bias one’s cognition to favor the in-group (often envisioned as a product of evolution). In contrast, other theories suggest that group-biases can emerge from nonspecialized cognitive processes. These perspectives have historically been difficult to disambiguate given that observed behavior can often be attributed to innate processes, even when groups are experimentally assigned. Here, we use modern techniques from the field of AI that allow us to ask what group biases can be expected from a learning agent that is a pure blank slate without any intrinsic social biases, and whose lifetime of experiences can be tightly controlled. This is possible because deep reinforcement-learning agents learn to convert raw sensory input (i.e. pixels) to reward-driven action, a unique feature among cognitive models. We find that blank slate agents do develop group biases based on arbitrary group differences (i.e. color). We show that the bias develops as a result of familiarity of experience and depends on the visual patterns becoming associated with reward through interaction. The bias artificial agents display is not a static reflection of the bias in their stream of experiences. In this minimal environment, the bias can be overcome given enough positive experiences, although unlearning the bias takes longer than acquiring it. Further, we show how this style of tabula rasa group behavior model can be used to test fine-grained predictions of psychological theories.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Google DeepMind, London EC4A 3TW, United Kingdom.
- Department of Psychology, University of Toronto, Toronto, ON M5S 3G3, Canada.
- Vector Institute, Toronto, ON M5G 1M1, Canada.
- Schwartz Reisman Institute for Technology and Society, University of Toronto, Toronto, ON M5G 1L7, Canada.
ORCID-Profile: Raphael Köster, Edgar A Duéñez-Guzmán, Joel Z Leibo
Zitieren
APA
Köster, R., Duéñez-Guzmán, E. A., Cunningham, W. A., & Leibo, J. Z. (2025). Tabula rasa agents display emergent in-group behavior. Proceedings of the National Academy of Sciences, 122(25). https://doi.org/10.1073/pnas.2319947121
Vancouver
Köster R, Duéñez-Guzmán EA, Cunningham WA, Leibo JZ. Tabula rasa agents display emergent in-group behavior. Proceedings of the National Academy of Sciences. 2025;122(25). doi:10.1073/pnas.2319947121
BibTeX
@article{Kster2025,
author = {Köster, Raphael and Duéñez-Guzmán, Edgar A and Cunningham, William A and Leibo, Joel Z},
title = {Tabula rasa agents display emergent in-group behavior},
journal = {Proceedings of the National Academy of Sciences},
year = {2025},
volume = {122},
number = {25},
doi = {10.1073/pnas.2319947121}
}
EUVIMED