Generalized Discrete Diffusion from Snapshots
This work addresses the challenge of efficient and high-quality discrete generation for machine learning applications, representing a significant advancement rather than an incremental improvement.
The authors tackled the problem of discrete diffusion modeling by introducing a unified framework that supports arbitrary noising processes over large discrete state spaces, resulting in outperforming existing discrete diffusion methods in training efficiency and generation quality and beating autoregressive models for the first time at this scale.
We introduce Generalized Discrete Diffusion from Snapshots (GDDS), a unified framework for discrete diffusion modeling that supports arbitrary noising processes over large discrete state spaces. Our formulation encompasses all existing discrete diffusion approaches, while allowing significantly greater flexibility in the choice of corruption dynamics. The forward noising process relies on uniformization and enables fast arbitrary corruption. For the reverse process, we derive a simple evidence lower bound (ELBO) based on snapshot latents, instead of the entire noising path, that allows efficient training of standard generative modeling architectures with clear probabilistic interpretation. Our experiments on large-vocabulary discrete generation tasks suggest that the proposed framework outperforms existing discrete diffusion methods in terms of training efficiency and generation quality, and beats autoregressive models for the first time at this scale. We provide the code along with a blog post on the project page : \href{https://oussamazekri.fr/gdds}{https://oussamazekri.fr/gdds}.