---
license: mit
pipeline_tag: text-generation
---

# ELF: Embedded Language Flows

ELF is a class of continuous diffusion language models based on continuous-time Flow Matching. Unlike existing diffusion language models (DLMs), ELF predominantly stays within the continuous embedding space until the final time step, where it maps to discrete tokens using a shared-weight network. This formulation makes it straightforward to adapt established techniques from image-domain diffusion models, such as classifier-free guidance (CFG).

- **Paper:** [ELF: Embedded Language Flows](https://huggingface.co/papers/2605.10938)
- **Repository:** [GitHub](https://github.com/lillian039/ELF)

## Evaluation and Inference

The official implementation is provided in JAX. To evaluate the pre-trained checkpoints, follow the installation instructions in the [GitHub repository](https://github.com/lillian039/ELF) and use the following commands from the `src/` directory:

### Unconditional Generation (OpenWebText)
```bash
python eval.py \
    --config configs/training_configs/train_owt_ELF-B.yml \
    --checkpoint_path embedded-language-flows/ELF-B-owt
```

### Conditional Generation (Summarization/Translation)
```bash
# XSum (summarization)
python eval.py \
    --config configs/training_configs/train_xsum_ELF-B.yml \
    --checkpoint_path embedded-language-flows/ELF-B-xsum

# WMT14 De-En (translation)
python eval.py \
    --config configs/training_configs/train_de-en_ELF-B.yml \
    --checkpoint_path embedded-language-flows/ELF-B-de-en
```

## Citation

```bibtex
@article{elf2026,
  title={ELF: Embedded Language Flows},
  author={Hu, Keya and Qiu, Linlu and Lu, Yiyang and Zhao, Hanhong and Li, Tianhong and Kim, Yoon and Andreas, Jacob and He, Kaiming},
  journal={arXiv preprint arXiv:2605.10938},
  year={2026}
}
```