Reference library
Large Language Diffusion Models
Paper summary
LLaDA uses masked diffusion rather than left-to-right factorization for language modeling. A Transformer predicts masked tokens during reverse generation, with pretraining and instruction-tuning evaluations comparing it to autoregressive baselines.
Relevance to the project
An alternative generative-process reference showing that language capability is not tied to one decoding convention. It remains peripheral to current neural-engine design.
Iterative denoising is externally specified inference, not self-originating behavioral change.
Source record
- Reference ID
- llada
- Archived version
- 2502.09992v3
- Summary provenance
- reference-library/papers/llada/metadata.json · summary
- Review status
- initial orientation summary; not a critical full-paper review
- Reading scope
- Selected abstract/opening and available closing discussion/conclusion passages in the preserved text extract; broader project notes provide context. No reproduction of experiments, exhaustive methods review, or equation/figure audit.