Cortex-wide representational drift of different layers
Pollak, Y. E.; Sachdev, R.; Larkum, M.; Gilad, A.
Show abstract
Representational drift, the gradual evolution of neural population codes over days, has been widely documented at the level of single neurons. However, whether drift is organized across larger cortical populations and spatial scales remains unclear. Here, we examined cortex-wide population dynamics under tightly controlled sensory input and behavior. Using widefield calcium imaging, we longitudinally recorded excitatory activity in Layers 2/3 (L2/3) or Layer 5 (L5) across 25 cortical areas while mice performed a whisker-based texture discrimination task. Sensory-evoked activity was tracked over five consecutive days during stable task performance. Population activity patterns reorganized over days, across the cortex, in both layers. This reorganization followed distinct laminar motifs: notably L5 responses exhibited a widespread, monotonic decrease in activity across most cortical areas, whereas L2/3 responses showed spatially localized and heterogeneous changes that were strongest during the sensory period. These laminar differences extended beyond the stimulus period, with L5 exhibiting more prolonged temporal engagement than L2/3. Together, these findings indicate that representational drift can unfold as a process that is coordinated across cortex with distinct laminar profiles. Thus, drift may reflect a structured feature of cortex-wide circuit dynamics.
Matching journals
The top 3 journals account for 50% of the predicted probability mass.
Similar papers in this journal
Similar papers in this journal
- Stimulus information guides the emergence of behavior related signals in primary somatosensory cortex during learning 98%
- NMDAR-dependent emergence of behavioral representation in primary visual cortex 97%
- Inferring functional organization of posterior parietal cortex circuitry based on information flow 97%
Similar papers in this journal
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.