Gamma oscillations have been implicated in many cognitive functions. Fast spiking interneurons are thought to play an important role in gamma synchrony. Recently, fast spiking interneurons in the entorhinal cortex have been shown to exhibit type 2 excitability and postinhibitory rebound (PIR). Theoretical work has shown that these properties make interneuronal network gamma (ING) more robust than in networks of type 1 interneurons. Here we show that this robust ING persists in a sparsely connected excitatory network. We also show that phase response curve (PRC) theory can predict under what circumstances the interneurons will sparsely synchronize in two clusters, and how increasing the delay and/or the conductance destabilizes two clusters in favor of a single cluster.