DAEDAL
[ICLR 2026] Official repository of "Beyond Fixed: Training-Free Variable-Length Denoising for Diffusion Large Language Models"
- 176
- Python
- Apache-2.0
- 9 mo ago
[ICLR 2026] Official repository of "Beyond Fixed: Training-Free Variable-Length Denoising for Diffusion Large Language Models"
A 4B diffusion grounding model for fast, precise visual evidence extraction. Unifies diverse grounding tasks through parallel decoding, with optional self-speculative decoding delivering 4.51x autoregressive throughput on COCO.
Embedding Inversion via Conditional Masked Diffusion: recover original text from embedding vectors using parallel denoising. Live demo + training pipeline + technical report.
Minimal masked diffusion LM in PyTorch. Pretraining + SFT + a scaling family from 50M to 350M, LLaDA recipe.
Diffusion model that writes renderable UI screens
Runs in the browser on Hugging Face
Diffusion model that writes renderable UI screens
Runs in the browser on Hugging Face