On The Plurality of Graphs
This addresses the problem of understanding communication dynamics in multi-agent AI systems, but appears incremental as it confirms a prior hypothesis.
The paper investigates how graph structure affects language emergence in multi-agent systems, showing that graph-generating processes and centrality-based edge sampling significantly influence communication dynamics.
We conduct a series of experiments designed to empirically demonstrate the effects of varying the structural features of a multi-agent emergent communication game framework. Specifically, we model the interactions (edges) between individual agents (nodes)as the structure of a graph generated according to a series of known random graph generating algorithms. Confirming the hypothesis proposed in [10], we show that the two factors of variation induced in this work, namely 1) the graph-generating process and 2) the centrality measure according to which edges are sampled, in fact play a significant role in determining the dynamics of language emergence within the population at hand.