<ul><li>Humans have a remarkable ability to disentangle complex sensory inputs into simple factors of variation without much supervision.</li><li>Existing works on disentanglement require a lot of auxiliary information, which may not provide conditional independence over the factors of variation.</li><li>This research explores how to achieve disentanglement when the auxiliary information does not ensure conditional independence.</li><li>The study shows theoretically and experimentally that disentanglement is possible with reduced auxiliary information dimension.</li></ul>

Towards efficient representation identification in supervised learning

Discover more