Abstract In practice, there exists numerous examples indicating the phenomenon that individuals are inclined to employ different strategies when interacting with the others. This is incurred by the fact that individuals are able to adjust their strategies adaptively under different interacting environments. Scholars have devoted their efforts to this topic; nevertheless, the impact of diverse interactions on cooperation still needs to be further explored. In this manuscript, we propose a mechanism aiming to investigate the co-evolution of personal reputation and strategy under a general framework of interactive diversity (being referred to as edge-strategy for simplicity). Numerous simulations are conducted with sufficient analyses of the obtained results being provided. As illustrated by the evolutionary dynamics, we find that there exists an optimal reputation value which can promote the frequency of cooperation by a large extent. Furthermore, we can clearly conclude that the consideration of interaction diversity is able to ensure the maintenance of cooperation even if the temptation to adopt antisocial behavior is relatively large. Aiming to understand the phenomenon better, we also quantitatively analyze the results by investigating the statistics of interaction chain, cluster size and other microscopic information. Overall, we hope the findings here can provide some interesting insights in solving social dilemmas.

中文翻译：

---

囚徒困境博弈中节点声誉与边缘策略的协同演化研究

摘要 在实践中，存在大量个体在与他人互动时倾向于采用不同策略的现象。这是因为个人能够在不同的交互环境下自适应地调整他们的策略。学者们致力于这个课题；尽管如此，各种互动对合作的影响仍有待进一步探讨。在这份手稿中，我们提出了一种机制，旨在研究在交互多样性的一般框架下个人声誉和策略的共同进化（为简单起见，称为边缘策略）。进行了大量模拟，并提供了对所获得结果的充分分析。正如进化动力学所说明的那样，我们发现存在一个最优的声誉值，可以在很大程度上促进合作的频率。此外，我们可以清楚地得出结论，即使采取反社会行为的诱惑相对较大，对互动多样性的考虑也能够确保合作的维持。为了更好地理解这一现象，我们还通过调查交互链、集群大小等微观信息的统计数据对结果进行了定量分析。总的来说，我们希望这里的发现可以为解决社会困境提供一些有趣的见解。我们可以清楚地得出结论，即使采取反社会行为的诱惑比较大，对互动多样性的考虑也能够确保合作的维持。为了更好地理解这一现象，我们还通过调查交互链、集群大小等微观信息的统计数据对结果进行了定量分析。总的来说，我们希望这里的发现可以为解决社会困境提供一些有趣的见解。我们可以清楚地得出结论，即使采取反社会行为的诱惑比较大，对互动多样性的考虑也能够确保合作的维持。为了更好地理解这一现象，我们还通过调查交互链、集群大小等微观信息的统计数据对结果进行了定量分析。总的来说，我们希望这里的发现可以为解决社会困境提供一些有趣的见解。