DiffusER: Discrete Diffusion via Edit-based Reconstruction
This addresses the limitation of current text generation models for users needing to revise or condition on prototypes, though it is incremental as it adapts existing diffusion methods to text.
The paper tackles the problem of text generation models lacking revision capabilities by introducing DiffusER, an edit-based generative model based on denoising diffusion, which rivals autoregressive models on tasks like machine translation, summarization, and style transfer.
In text generation, models that generate text from scratch one token at a time are currently the dominant paradigm. Despite being performant, these models lack the ability to revise existing text, which limits their usability in many practical scenarios. We look to address this, with DiffusER (Diffusion via Edit-based Reconstruction), a new edit-based generative model for text based on denoising diffusion models -- a class of models that use a Markov chain of denoising steps to incrementally generate data. DiffusER is not only a strong generative model in general, rivalling autoregressive models on several tasks spanning machine translation, summarization, and style transfer; it can also perform other varieties of generation that standard autoregressive models are not well-suited for. For instance, we demonstrate that DiffusER makes it possible for a user to condition generation on a prototype, or an incomplete sequence, and continue revising based on previous edit steps.