Innovation networks play a key role in advancing knowledge transfer, collaboration, and technological progress across sectors and regions. Central to the understanding of the mechanisms driving such networks is their dynamic evolution and structure. Much of the literature explores spatial and socio-economic drivers of innovation networks, focusing on geographic, institutional, and cultural influences. However, many of these studies tend to overlook the intricate properties that govern the behavior and dynamics of these networks. This study seeks to address this research gap, delving deeper by investigating Chinese intercity innovation networks between 2007 and 2018. Specifically, we examine: (i) the preferential attachment dynamics within intercity innovation networks, (ii) transitivity effects that underscore the interconnectedness of these networks, and (iii) the persistence and recurrence of connections. We find that cities show indeed a remarkable tendency to form ties with others that already have numerous connections. Such transitivity effects are important in highlighting the formation of innovation clusters. Moreover, the influence of link memory suggests that past collaborations significantly determine future partnerships, similar to the persistent nature of relationships in agglomeration theories.