So to put things simply, this comes from a view of a neuron (or network of neurons) as a dynamical system with various states. These states are constantly in flux and we can describe the states as patterns that are seen from plotting the pieces of the instantaneous state of the system in question. These patterns will repeat and change over time based on inputs and such, hence the term "phase change"

At least that's my current understanding of the matter, coming from the mouth of an undergrad late in a biochem degree with a focus on neuro