ShortListing Model: A Streamlined Simplex Diffusion for Biological Sequence Generation
Abstract
Generative modeling of discrete variables is challenging yet crucial for applications in natural language processing and biological sequence design. We introduce the Shortlisting Model (SLM), a novel simplex-based diffusion model inspired by progressive candidate pruning. SLM operates on simplex centroids, reducing complexity and enhancing scalability. Additionally, SLM incorporates a flexible implementation of classifier-free guidance, enhancing unconditional generation performance. Extensive experiments in DNA promoter and enhancer design, and protein design demonstrate SLM's competitive performance and scalability.
Cite
Text
Song et al. "ShortListing Model: A Streamlined Simplex Diffusion for Biological Sequence Generation." ICLR 2025 Workshops: MLGenX, 2025.Markdown
[Song et al. "ShortListing Model: A Streamlined Simplex Diffusion for Biological Sequence Generation." ICLR 2025 Workshops: MLGenX, 2025.](https://mlanthology.org/iclrw/2025/song2025iclrw-shortlisting/)BibTeX
@inproceedings{song2025iclrw-shortlisting,
title = {{ShortListing Model: A Streamlined Simplex Diffusion for Biological Sequence Generation}},
author = {Song, Yuxuan and Zhang, Zhe and Pei, Yu and Gong, Jingjing and Wang, Mingxuan and Zhou, Hao and Liu, Jingjing and Ma, Wei-Ying},
booktitle = {ICLR 2025 Workshops: MLGenX},
year = {2025},
url = {https://mlanthology.org/iclrw/2025/song2025iclrw-shortlisting/}
}