Ask HN: What Happened to dLLMs?

Diffusion LLMs (dLLMs) seemed so promising a year ago, but I don't see much about them now. What happened?

3 points | by ray__ 1 day ago

4 comments

  • simulacra8 2 hours ago
    dLLMs are still promising. It's a counterintuitive hardware limitation. You should watch this: https://www.youtube.com/watch?v=r305-aQTaU0
  • acangiano 1 day ago
    Because they shipped and are better only for specific applications. For everything else, they lag behind frontier models. They win in latency-sensitive niches (e.g., autocomplete, code infilling, subagent calls). Not frontier-level reasoning.
  • diffusionmodels 1 day ago
    They’re kinda dyslexic - like bad OCR
  • tmpsvc2695f5 22 hours ago
    [dead]