HIGHLIGHTS
- who: Jorge Fernandez-de-Cossio-Diaz and collaborators from the LaboratorySorbonne Universitu00e9, Paris, France have published the research: Disentangling Representations in Restricted Boltzmann Machines without Adversaries, in the Journal: (JOURNAL) of 21/07/2022
- what: The authors propose a simple effective way of disentangling representations without any need to train adversarial discriminators and apply the approach to Restricted Boltzmann Machines one of the simplest representation-based generative models. The authors show how the framework allows for analytically computing the cost in terms of the loglikelihood of the data associated with the disentanglement of their representations . . .
If you want to have access to all the content you need to log in!
Thanks :)
If you don't have an account, you can create one here.