2025/06/25 by Pang, Jiahuan, Wang, Wendong
#Adaptation and Self-Organizing Systems (nlin.AO) #FOS: Physical sciences #Statistical Mechanics (cond-mat.stat-mech)
paper · doi:10.48550/arxiv.2506.20888
Understanding the flow of information among individuals in a complex system is critical to our understanding of its dynamics. The concept of influence is often used to correlate with information transfer, as quantified by various information-theoretical tools. However, influence has not been explicitly and quantitatively defined. Here, we propose an intuitive and quantifiable definition of influence and use the Vicsek model as an example to investigate the quantitative relations between influence and information transfer. We identify quantitative relations between influence and information transfer in pairwise interactions and collective interactions. Significantly, we uncover the dual effect of noise. Finally, we use partial information decomposition methods to reveal the unique and synergistic nature of influence and information transfer in a collective. The insights gained in our model will contribute to our understanding of other natural and artificial complex systems.