Skip to content
anythingai.space
Inference & Servingadvanced12 min · 9 steps

Diffusion Language Models

Instead of writing left to right, start from a fully masked sequence and denoise every position in parallel — fast, good at code and infill, still behind on hard reasoning.

forward passes8
tokens / pass1
GENERATED SO FARdefpass 1add(a,pass 2b):pass 3returnpass 4[MASK]pass 5[MASK]pass 6[MASK]pass 7[MASK]pass 8AR modelleft → right · past only
token
metrics are illustrative
1 / 9
ProblemStep 1 of 9

Autoregressive: one token per forward pass

Almost every LLM you use is autoregressive (AR): it predicts each token from the past only, left to right, one forward pass per token.

Eight tokens means eight sequential passes, no matter how predictable the text is.

Lesson 8 of 9 in Inference & Serving