Diffusing Gaussian Mixtures for Generating Categorical Data
This work addresses the problem of generating categorical data for applications like protein modeling, but it is incremental as it builds on existing diffusion models with a focus on sample quality rather than fundamental novelty.
The authors tackled the challenge of generating categorical data by proposing a diffusion-based generative model that operates in a continuous domain while incorporating categorical structure, resulting in high-quality samples evaluated on synthetic and real-world protein datasets.
Learning a categorical distribution comes with its own set of challenges. A successful approach taken by state-of-the-art works is to cast the problem in a continuous domain to take advantage of the impressive performance of the generative models for continuous data. Amongst them are the recently emerging diffusion probabilistic models, which have the observed advantage of generating high-quality samples. Recent advances for categorical generative models have focused on log likelihood improvements. In this work, we propose a generative model for categorical data based on diffusion models with a focus on high-quality sample generation, and propose sampled-based evaluation methods. The efficacy of our method stems from performing diffusion in the continuous domain while having its parameterization informed by the structure of the categorical nature of the target distribution. Our method of evaluation highlights the capabilities and limitations of different generative models for generating categorical data, and includes experiments on synthetic and real-world protein datasets.