KERMIT: Generative Insertion-Based Modeling for Sequences

William ChanNikita KitaevKelvin GuuMitchell SternJakob Uszkoreit

   Papers with code   Abstract  PDF

We present KERMIT, a simple insertion-based approach to generative modeling for sequences and sequence pairs. KERMIT models the joint distribution and its decompositions (i.e., marginals and conditionals) using a single neural network and, unlike much prior work, does not rely on a prespecified factorization of the data distribution... (read more)

Benchmarked Models

No benchmarked models yet. Click here to submit a model.