When to Plan, When to Polish: Noise Level as a Granularity Axis for Diffusion Language Models

Excerpt

arXiv:2606.21802v2 Announce Type: replace Abstract: Standard tokenwise diffusion LMs keep training corruption and inference commitment at token granularity throughout denoising. At high noise, this leaves scattered local fragments rather than coherent evidence, making it hard to form early coarse structure, exactly what planning-sensitive generation requires. Hierarchical planning methods add coarse stages to separate planning from wording, but they need extra planners, block latents, or two sta