Diffusion language models (DLMs) enable parallel token generation, potentially overcoming the sequential limitations of autoregressive (AR) decoding. However, DLMs currently underperform AR models in quality due to a lack of introspective consistency, where AR models align with their generated outputs.

Speculative decoding accelerates autoregressive inference by using a fast draft model to predict upcoming tokens from a slower target model. It verifies predictions in parallel with a single forward pass of the target model, addressing the sequential dependency bottleneck.