Understanding memory encoding and recall processes in the hippocampus is pivotal for unravelling brain function. We employed two-photon calcium imaging to observe neuronal activity (hippocampal CA1 neurons) in behaving mice performing spatial memory tasks, capturing the dynamics of memory-related circuits. Shannon Information Theory provides a useful toolbox, applicable to the analysis of calcium imaging datasets, to assay changes in information representation and processing in brain circuits while mice are performing such cognitive tasks. In particular, the recently developed Partial Information Decomposition (PID) framework, by facilitating the dissection of neural data into “atoms” of redundancy, synergy, unique information, and mutual information, can shed light on how neurons collectively process information. We used the Kraskov mutual information estimator, using the approach of Bertschinger et al (Entropy 16:2161-83, 2014) to quantify the unique information shared by pairs of neurons, and thus decompose the mutual information into components reflecting synergistic and/or redundant interactions between neurons. Our analysis reveals a decorrelation in neural networks upon encountering new environments, suggesting a reliance on synergistic information for memory encoding. This methodology not only deepens our comprehension of hippocampal function but also paves the way for explaining the effects on information processing of abnormal network activity in neurodegenerative diseases. Partial Information Decomposition offers novel insights into normal and pathological changes in memory processing, with the potential to contribute to the investigation of neurodegenerative disorders such as Alzheimer’s disease.