AI RESEARCH
CRoCoDiL: Continuous and Robust Conditioned Diffusion for Language
arXiv CS.AI
•
ArXi:2603.20210v1 Announce Type: cross Masked Diffusion Models (MDMs) provide an efficient non-causal alternative to autoregressive generation but often struggle with token dependencies and semantic incoherence due to their reliance on discrete marginal distributions. We address these limitations by shifting the diffusion process into a continuous sentence-level semantic space.