Early-Signal Detector

model · first seen 4 Oct 2026

diffusion language model

Also: DLM, DLLM, MDLM, diffusion LLM, masked diffusion, diffusion adapters, discrete diffusion, masked diffusion LM

Language model that generates text by iteratively denoising masked or noisy token sequences.

Diffusion language models generate text by starting from a masked or noisy token sequence and refining it over several denoising steps, rather than producing one token at a time from left to right. Most of the sources are arXiv papers that build on open models such as LLaDA and Dream. They address faster decoding through speculative branches and attention pruning, better-calibrated choice of which tokens to commit, remasking and self-correction, and post-training with ELBO surrogates. Practical activity is appearing too: a Mojo repo for local CPU inference, a 35B mixture-of-experts diffusion model aimed at agentic coding, and a Hacker News thread on Inception Labs' Mercury 3.

If the remaining problems are solved (inference cost per step, miscalibrated confidence, and decoding that drifts back to pseudo-autoregressive order), parallel generation could give a real latency and throughput alternative to autoregressive LLMs, especially for code. The spread from papers into released weights and local inference tooling suggests the approach is moving beyond pure research.

Analyst note (emerging), 17 Nov 2026.

Mentions per week

04119 May11 Aug10 NovPapersCodeDiscussion

PapersCodeDiscussion

Rank 7 on 16 Nov 2026 · 72 mentions in 28 days · 69 independent groups · papers → code → discussion · first seen 6 weeks ago · acceleration -0.03

Earliest sources

Latest sources

Mentioned together with

parallel decodingspeculative decodingKV cachetest-time scalinggroup relative policy optimizationon-policy distillationknowledge distillationsparse attentionmixture of expertsfull-duplex speech model

In the ledger