Rethinking Soft Tokens for Parallel Decoding in Diffusion Language Models
Diffusion language models (DLMs) enable parallel generation by predicting and committing multiple tokens at each denoising step, yet they can generate individually plausible but mutually inconsistent tokens. Recent work shows that \emph{soft tokens} can mitigate this issue by representing uncertain positions with conti...
Kodai Kawamura, Kenji Kawaguchi, Anji Liu
· 0 citations