In social dilemmas, individuals need to forgo short-term temptations to achieve synergistic collective outcomes through cooperation. Previous work has examined mechanisms through which cooperation can evolve, including direct reciprocity, indirect reciprocity, environmental stochasticity, network reciprocity, and demographic stochasticity. These mechanisms have largely been studied under natural selection or social learning. As the other side of the same coin, it is equally important to study cooperation under self-learning, where individuals adapt based on their own experiences. Here, we focus on the dynamics of multi-agent reinforcement learning and derive analytical conditions under which these mechanisms can stabilize cooperation under learning dynamics across scales. We find that reinforcement learning can steer self-interested individuals toward cooperation when agents value the future over short-term temptation. Our work provides a unified approach to identifying the determinants of learning to cooperate in a changing world, thereby paving the way for the advancement of cooperative artificial intelligence.
The results reveal that reputation-modulated learning significantly promotes the emergence of cooperative behavior, and the discontinuous phase transition from full cooperation to full defection as the temptation increases.
Chenyang Zhao, Ji-Qiang Zhang, Li Chen et al.· 0 citations
Mechanistic analyses show that a moderate neighborhood size enables individuals to strike an optimal balance between information sufficiency and decision-making tractability, which allows them to detect reciprocal opportunities while avoiding the deterioration of decision quality due to information overload.
Yi-Hsin Ku, Xin Ou, Ji-Qiang Zhang et al.· 0 citations
Human cooperation unfolds in social environments where individuals influence each other through payoff-based learning and social comparison, the tendency to evaluate fitness relative to others. However, it is still unclear how social comparison and population structure jointly shape cooperation. Here, we incorporate so...
Xiao-Jin Xiong, Qin Li, J. Kurths et al.· Communications Physics· 0 citations
An evaluation framework that measures agents' empirical and normative expectations in addition to behavioral convergence is introduced, offering designers of multi-agent systems a principled basis for selecting the social processes that sustain cooperation.
Various theoretical and empirical studies have accounted for why humans cooperate in competitive environments. Although prior work has revealed that network structure and multiplex interactions can promote cooperation, most theory assumes that individuals play dilemma games in all social contexts. However, real-world a...