Adaptive stretching of representations across brain regions and deep learning model layers

Xin-Ya Zhang, Sebastian Bobadilla-Suarez, Xiaoliang Luo, Marilena Lemonari, Scott L. Brincat, Markus Siegel, Earl K. Miller, Bradley C. Love. bioRxiv.

This code repository includes the following documents

data: cue images and electrode monkey spike data sample from [1];

[1] Siegel, M., Buschman, T. J. & Miller, E. K. Cortical information flow during flexible sensorimotor decisions. Science 348, 1352–1355 (2015).

script: neural data and model data processing, including:

*neuralcode: process neuronal data from monkey spike data and get representational dissimilarity matrix (RDM);

*NNmodel: the CNN-LSTM model example was fed the same movies as that presented to monkeys in the monkey experiments and closely replicated their training procedure;

*demo: Example codes for plotting the data figures in the main text.

Software Dependencies

• Python (>= 3.8)
• NumPy (>= 1.21.2)
• Pytorch (>= 1.10.0)
• Pyspike (=0.8.0)
• Scikit-learn (>=1.0.2)
• Scipy (>= 1.7.3)
• Seaborn (>= 0.11.2)
• Pandas (>= 1.4.1)
• Matplotlib (>= 3.5.1)
• Torch-intermediate-layer-getter (0.1.post1)