Humans flexibly integrate social information despite interindividual differences in reward DOI Creative Commons
Alexandra Witt, Wataru Toyokawa, Kevin N. Laland

et al.

Proceedings of the National Academy of Sciences, Journal Year: 2024, Volume and Issue: 121(39)

Published: Sept. 20, 2024

There has been much progress in understanding human social learning, including recent studies integrating information into the reinforcement learning framework. Yet previous often assume identical payoffs between observer and demonstrator, overlooking diversity of real-world interactions. We address this gap by introducing a socially correlated bandit task that accommodates payoff differences among participants, allowing for study under more realistic conditions. Our Social Generalization (SG) model, tested through evolutionary simulations two online experiments, outperforms existing models incorporating generalization process, but treating it as noisier than individual observations. findings suggest is flexible previously believed, with SG model indicating potential resource-rational trade-off where partially replaces exploration. This research highlights flexibility humans’ us to integrate from others different preferences, skills, or goals.

Language: Английский

Visual-spatial dynamics drive adaptive social learning in immersive environments DOI Open Access
Charley M. Wu, Dominik Deffner,

Benjamin Kahl

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2023, Volume and Issue: unknown

Published: June 29, 2023

Human cognition is distinguished by our ability to adapt different environments and circumstances. Yet the mechanisms driving adaptive behavior have predominantly been studied in separate asocial social contexts, with an integrated framework remaining elusive. Here, we use a collective foraging task virtual Minecraft environment integrate these two fields, leveraging automated transcriptions of visual field data combined high-resolution spatial trajectories. Our behavioral analyses capture both structure temporal dynamics interactions, which are then directly tested using computational models sequentially predicting each decision. These results reveal that adaptation selective learning driven individual success (rather than factors). Furthermore, it degree adaptivity---of learning---that best predicts performance. findings not only theories across domains, but also provide key insights into adaptability human decision-making complex dynamic landscapes.

Language: Английский

Citations

12
Humans flexibly integrate social information despite interindividual differences in reward DOI Creative Commons
Alexandra Witt, Wataru Toyokawa, Kevin N. Laland

et al.

Proceedings of the National Academy of Sciences, Journal Year: 2024, Volume and Issue: 121(39)

Published: Sept. 20, 2024

There has been much progress in understanding human social learning, including recent studies integrating information into the reinforcement learning framework. Yet previous often assume identical payoffs between observer and demonstrator, overlooking diversity of real-world interactions. We address this gap by introducing a socially correlated bandit task that accommodates payoff differences among participants, allowing for study under more realistic conditions. Our Social Generalization (SG) model, tested through evolutionary simulations two online experiments, outperforms existing models incorporating generalization process, but treating it as noisier than individual observations. findings suggest is flexible previously believed, with SG model indicating potential resource-rational trade-off where partially replaces exploration. This research highlights flexibility humans’ us to integrate from others different preferences, skills, or goals.

Language: Английский

Citations

3