Research Article

Discovering and Characterizing Hidden Variables Using a Novel Neural Network Architecture: LO-Net

Figure 7

Simulated robot data: outputs from the L-nets.
193146.fig.007a
(a) Output from L-net 1 for predicting 𝑠 𝑡 + 1 in 2-LO-net architecture ( 𝑧 and 𝑆 are the hidden variables here)
193146.fig.007b
(b) Output from L-net 2 for predicting 𝑧 𝑡 + 1 in 2-Lo-net architecture ( 𝑧 and 𝑆 are hidden variables here)