Using Canonical Polyadic (CP) Decomposition of Neural Tensors to Explore the Existence of a Feature-Centric, Rather Than Stimulus-Specific, Learning Signal in the Visual Cortex

Neuroscience data is inherently multiway (neurons × time × trials), yet standard pipelines flatten them and discard separable structure. We adopt a tensor-native workflow for cortical population analysis that preserves this organization using Canonical Polyadic (CP/PARAFAC) decomposition and CP-structured regression. Applied to calcium-imaging recordings from mouse visual cortex, CP factorization yields interpretable neuron, time, and trial components without vectorization. We use this framework to investigate a feature-centric hypothesis of plasticity—namely, that the learning drive acts at the level of neuronal features (cells/components) rather than stimulus labels. It carries no stimulus identity; instead, it selects and strengthens neurons that best support the learned representation while weakening others. Concretely, we ask whether reactivations associated with distinct visual stimuli reflect a neuron-level learning signal that jointly drives potentiation and depotentiation across stimulus-specific pathways. Our findings highlight how CP decomposition methods, although still uncommon in practice, offer interpretability, enhance statistical efficiency, and enable hypothesis-driven analyses of neural populations.