Understanding the complexity within neural circuits is crucial for unraveling the flexibility and the efficiency of the biological brain. In this study, we delved into the assessment of signal complexity among neuron populations within the somatosensory cortex (S1). Our approach aimed to capture population-level signal intricacies by employing highly repeatable and resolvable visual, tactile, and visuo-tactile inputs, coupled with the recording of neuronal unit activity at a remarkably high temporal resolution. Our findings unveiled an exceptionally high-dimensional state space characterising the spontaneous activity within S1 populations. Particularly noteworthy was the profound modulation of tactile input processing by visual inputs, showcasing the separation of even subtle nuances within visual input patterns. Furthermore, the dynamic activity states of the S1 neuron population persisted long after the cessation of stimulation, revealing resident information that could potentially serve as a substrate for a working memory. The recorded high-dimensional representations not only conveyed rich multimodal signals but also exhibited characteristics reminiscent of internal working memory-like signals. These findings shed light on the intricate and highly complex operations of cortical circuitry, emphasising the role of multimodal integration and the potential establishment of resident information contributing to cognitive processes within the somatosensory cortex. This study contributes valuable insights into the multifaceted nature of neural processing.