Therefore, this approach unifies the powerful and versatile notion of topics based on transformer embeddings with fully probabilistic modeling, as in models such as latent Dirichlet allocation (LDA).

The work I've done about the "Analysis of the latent space in the variational auto-encoder" to complete my Master's degree.

Abstract: Aiming at learning a probabilistic distribution over data, generative models have been actively studied with broad applications. This paper proposes a complex recurrent variational ...