图像生成被扩散模型统治，文本生成却始终是自回归的天下——文本能不能也用「从噪声里逐步去噪」的方式生成？扩散语言模型（diffusion LM）从学术好奇变成开源可用（LLaDA 等），双向生成与可修改性展示了独特优势，质量差距与效率难题也依然真实。本篇拆解它的机制、与自回归的三点对比、以及这场范式之争的现状。

Diffusion large language models (dLLMs) achieve high decoding efficiency through block-parallel, arbitrary-order generat…

Masked diffusion language models generate tokens in parallel, but their efficiency depends on which predictions are comm…

Masked diffusion language models (dLLMs) generate text by iteratively denoising masked tokens with bidirectional attenti…