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

<table>FVDS (three-LO-net architecture): comparing output from L-net two and hidden variable <svg height="9.8625002" id="M189" style="vertical-align:-2.29482pt" version="1.1" viewbox="0 0 7.875 9.8625002" width="7.875" xmlns="http://www.w3.org/2000/svg">
<g transform="matrix(1.25,0,0,-1.25,0,9.8625)">
<g transform="translate(72,-64.11)">
<text transform="matrix(1,0,0,-1,-71.95,66.45)">
<tspan style="font-size: 12.50px; " x="0" y="0">𝑦</tspan>
</text>
</g>
</g>
</svg>, correlation is 0.8964.</table>

Journal of Robotics

fig13

Figure 13

Figure 13: Discovering and Characterizing Hidden Variables Using a Novel Neural Network Architecture: LO-Net 