On the Role of Discreteness in Diffusion LLMs
This work addresses a fundamental problem for researchers in natural language processing and diffusion modeling, but it is incremental as it primarily analyzes existing approaches and motivates future improvements.
The paper tackles the challenge of applying diffusion models to language generation by analyzing how their mechanics conflict with the discrete and structured nature of text, identifying key issues such as uniform corruption and token-wise marginal training that hinder performance.
Diffusion models offer appealing properties for language generation, such as parallel decoding and iterative refinement, but the discrete and highly structured nature of text challenges the direct application of diffusion principles. In this paper, we revisit diffusion language modeling from the view of diffusion process and language modeling, and outline five properties that separate diffusion mechanics from language-specific requirements. We first categorize existing approaches into continuous diffusion in embedding space and discrete diffusion over tokens. We then show that each satisfies only part of the five essential properties and therefore reflects a structural trade-off. Through analyses of recent large diffusion language models, we identify two central issues: (i) uniform corruption does not respect how information is distributed across positions, and (ii) token-wise marginal training cannot capture multi-token dependencies during parallel decoding. These observations motivate diffusion processes that align more closely with the structure of text, and encourage future work toward more coherent diffusion language models.