GitHub - kylehkhsu/modular

This repo contains code for the paper "Range, not independence, drives modularity in biological representations.".

The following files can be used to generate each of the figures:

Figure 1c: Fig1_Basic.ipynb
Figure 1d: 'Figure 1d.ipynb'
Figure 2 Left: whatwherefigureplots.ipynb
Figure 2 Center and Right: autoencoding/
Figure 3a-d: RNN_linear_modularity_torch.py & RNN_nonlinear_modularity.py.
Figre 3e: see 'Actionable Neural Representations: Grid Cells from Minimal Constraints', 2023.
Figure 3f-g: 'Figure 3fg.ipynb'
Figure 4: Grid_Warp_Train.ipynb and Grid_Warp_Plot.ipynb
Figure 5: Correlated_Missing_Vars_Train.ipynb and Correlated_Missing_Vars_Plot.ipynb

Name		Name	Last commit message	Last commit date
Latest commit History 31 Commits
.idea		.idea
autoencoding		autoencoding
.DS_Store		.DS_Store
Correlated_Missing_Vars_Plot.ipynb		Correlated_Missing_Vars_Plot.ipynb
Correlated_Missing_Vars_Train.ipynb		Correlated_Missing_Vars_Train.ipynb
Dual_Motor_RNNs.ipynb		Dual_Motor_RNNs.ipynb
Fig1_Basic.ipynb		Fig1_Basic.ipynb
Fig1_Points.npy		Fig1_Points.npy
Figure 1d.ipynb		Figure 1d.ipynb
Figure 3fg.ipynb		Figure 3fg.ipynb
Grid_Warp_Plot.ipynb		Grid_Warp_Plot.ipynb
Grid_Warp_Train.ipynb		Grid_Warp_Train.ipynb
P&B_Plot.ipynb		P&B_Plot.ipynb
P&B_Train.ipynb		P&B_Train.ipynb
README.md		README.md
RNN_linear_modularity_torch.py		RNN_linear_modularity_torch.py
RNN_nonlinear_modularity.py		RNN_nonlinear_modularity.py
Single_Motor_RNNs.ipynb		Single_Motor_RNNs.ipynb
Warden_Plot.ipynb		Warden_Plot.ipynb
Warden_Train.ipynb		Warden_Train.ipynb
Xie_2022.py		Xie_2022.py
Xie_Plot.ipynb		Xie_Plot.ipynb
disentangled_rnn.py		disentangled_rnn.py
disentangled_rnn_params.py		disentangled_rnn_params.py
disentangled_rnn_train_clean.py		disentangled_rnn_train_clean.py
disentangled_rnn_utils.py		disentangled_rnn_utils.py
metrics.py		metrics.py
whatwherefigureplots.ipynb		whatwherefigureplots.ipynb

Provide feedback