What? Authors demonstrate the ability of DNNs to compress the training data and also measure how much current deep models actually compress data. Why? Compression…
Mostly my mind dump, sometimes useful stuff
What? Authors demonstrate the ability of DNNs to compress the training data and also measure how much current deep models actually compress data. Why? Compression…
What? This work proposes a method to learn the symmetries of a dynamical environment without supervision and prior information of such symmetries. Why? In such…
What? Models use an internal representation which encodes information about the state of the world. We would like such a representation to be robust to…