Propagation of chaos, which is commonly studied in the context of mean-field interacting particle systems, describes the asymptotic independence of particles in the infinite-population limit. It was recently shown by Lacker, Ramanan, and Wu, that random pairs of particles in strongly interacting particle systems on sparse graphs become asymptotically independent as well. In this talk I will discuss recent work on quantitative estimates for this phenomenon, which we refer to as unimodular propagation of chaos. The proof utilizes coupling estimates for random graphs as well as a surprising connection between propagation of chaos and first-passage percolation. This is joint work with Kavita Ramanan and Dan Lacker.