
Tanishq Mathew Abraham, Ph.D.
@iScienceLuvr · Nov 6, 2025
Diffusion Language Models are Super Data Learners
"when unique data is limited, diffusion language models (DLMs) consistently surpass autoregressive (AR) models by training for more epochs."
"At scale, a 1.7B DLM trained with a ∼ 1.5T - token compute budget on 10B unique

Elon Musk
@elonmusk
12:39 PM · November 7, 2025 · 53.7K views
86
43
510