Thanks, yes, I was thinking specifically of "Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution". They actually consider two noise distributions: one with uniform sampling for each noised token position, and one with a terminal masking (the Q^{uniform} and Q^{absorb}.) However, the terminal-masking system is clearly superior in their benchmarks.
1. Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution - https://arxiv.org/abs/2310.16834
2. Simple and Effective Masked Diffusion Language Models - https://arxiv.org/abs/2406.07524