Unifying Continuous and Discrete Text Diffusion with Non-simultaneous Diffusion Processes

TL;DR

NeoDiff unifies discrete and continuous text diffusion models using a novel Poisson process and adaptive denoising, leading to improved text generation quality and more flexible control over the diffusion process.

Contribution

The paper introduces NeoDiff, a new diffusion model that combines strengths of discrete and continuous approaches with a Poisson process and adaptive scheduling.

Findings

NeoDiff outperforms existing non-autoregressive diffusion models.

NeoDiff achieves superior quality on multiple text generation tasks.

The model demonstrates flexible and fine-grained control over diffusion processes.

Abstract

Diffusion models have emerged as a promising approach for text generation, with recent works falling into two main categories: discrete and continuous diffusion models. Discrete diffusion models apply token corruption independently using categorical distributions, allowing for different diffusion progress across tokens but lacking fine-grained control. Continuous diffusion models map tokens to continuous spaces and apply fine-grained noise, but the diffusion progress is uniform across tokens, limiting their ability to capture semantic nuances. To address these limitations, we propose \textbf{\underline{N}}on-simultan\textbf{\underline{e}}ous C\textbf{\underline{o}}ntinuous \textbf{\underline{Diff}}usion Models (NeoDiff), a novel diffusion model that integrates the strengths of both discrete and continuous approaches. NeoDiff introduces a Poisson diffusion process for the forward process, enabling a flexible and fine-grained noising paradigm, and employs a time predictor for the reverse process to adaptively modulate the denoising progress based on token semantics. Furthermore, NeoDiff utilizes an optimized schedule for inference to ensure more precise noise control and improved performance. Our approach unifies the theories of discrete and continuous diffusion models, offering a more principled and effective framework for text generation. Experimental results on several text generation tasks demonstrate NeoDiff's superior performance compared to baselines of non-autoregressive continuous and discrete diffusion models, iterative-based methods and autoregressive diffusion-based methods. These results highlight NeoDiff's potential as a powerful tool for generating high-quality text and advancing the field of diffusion-based text generation.