Don't Retrain, Align: Adapting Autoregressive LMs to Diffusion LMs via Representation Alignment

ArXi:2605.06885v1 Announce Type: cross Diffusion language models (DLMs) have recently nstrated capabilities that complement standard autoregressive (AR) models, particularly in non-sequential generation and bidirectional editing. Although recent work has shown that pretrained autoregressive checkpoints can be converted into diffusion language models, existing recipes primarily transfer parameters through continued denoising