To address the challenge of scarce computational resources in genomic modeling, we introduce GERM, a genomic foundation model with strong compression performance and fast adaptability. GERM improves upon models like DNABERT-2 by eliminating outliers that hinder low-rank adaptation and post-training quantization, enhancing both efficiency and robustness. We replace the vanilla attention layer with an outlier-free mechanism inspired by associative memory models. By removing outliers during both pre-training and fine-tuning, this approach accelerates adaptation, reduces computational costs, and enhances quantization robustness within acceptable loss margins. Additionally, we propose GERM-T, a strategy that employs small-step continual learning within the outlier-free framework, leveraging original checkpoints to avoid retraining from scratch. Empirically, GERM improves fine-tuning performance by 37.98% and quantization by 64.34% over the baseline model. It also reduces average kurtosis by 92.14% and maximum infinity norm by 82.77%. Compared to leading methods, GERM consistently delivers superior performance, offering a practical solution for genomic modeling in resource-constrained settings. Code is available atthis https URL.
View on arXiv@article{luo2025_2505.00598,
title={ Fast and Low-Cost Genomic Foundation Models via Outlier Removal },
author={ Haozheng Luo and Chenghao Qiu and Maojiang Su and Zhihan Zhou and Zoe Mehta and Guo Ye and Jerry Yao-Chieh Hu and Han Liu },
journal={arXiv preprint arXiv:2505.00598},
year={ 2025 }
}