Pooling neurons masks heterogeneous low-dimensional geometries during temporal processing in DMFC

Project Description

• This project examines whether neurons recorded within the same macaque brain region form functionally distinct subpopulations with heterogeneous population dynamics during behavior.

Objectives and Approach

• Identify distinct neural subpopulations and characterize their contributions to task-related computations using linear dimensionality reduction.

Progress

• Reproduced the main findings and figures from the paper associated with the dataset.Next Steps (Done)

• Unsupervised clustering identified two groups of neurons with distinct response profiles. (Done)

• Geometric differences between the two populations were observed in principal component space (Done)

• One of the populations distinguishes between the two prior conditions in principal component space (Done)

Background and References

• Sohn, H., Narain, D., Meirhaeghe, N., & Jazayeri, M. (2019). Bayesian computation through cortical latent dynamics. Neuron, 103(5), 934-947.