Synaptic and representational drift describe the dynamic changes in neuronal connections and activity over time. However, the relationship between these two processes remains elusive. Here, we integrated representational dynamics and synaptic drift into a firing rate model to explore the influence of synaptic drift on representational drift.In mouse auditory cortex, neuronal populations respond to stimuli in a clustered manner (Aschauer et al., 2022). Our model, comprising inhibitory and excitatory neurons, reproduces these response patterns, especially under inhibition-dominated conditions. We introduced synaptic drift as a gradual, random alteration of synaptic connectivity, mirroring experimental observations (Loewenstein et al., 2011). This led to two dynamical states in the network: periods of relative stability with slow ongoing changes in representations and occasional abrupt transitions to different activity patterns. The resulting bimodal distribution, also present in experimental data, suggests two distinct processes of response changes. We examined these continuous and abrupt response transitions in our network model by analyzing the numerically identified stable and unstable fixed points, using a method adapted from Sussilo and Barak (2013). We found two specific fixed point changes responsible for these two types of response transitions: slow ongoing representation changes corresponded to moving stable fixed points, whereas abrupt transitions coincided with emerging or disappearing unstable fixed points.We conclude that in a generic network model, random synaptic changes produce both gradual and abrupt response changes, similar to experimental findings. In our network model these qualitatively different response changes manifest through specific alterations of the network's fixed points.