GenoJEPA-Evaluation

This dataset provides the downstream evaluation benchmark used for GenoJEPA, a genomic representation learning framework based on joint-embedding predictive architecture.

The evaluation resource is designed to test the transferability of genomic sequence representations across diverse classification tasks. It supports both full fine-tuning and frozen-embedding probing protocols.

Contents

This resource contains downstream genomic sequence classification tasks in train/test format. Each task provides DNA sequences and labels for evaluating genomic foundation models.

The benchmark is used in GenoJEPA to evaluate fine-tuning performance, frozen embedding probing, and lightweight classifiers such as Logistic Regression.

For detailed task definitions, data layout, evaluation scripts, and usage examples, please refer to the GitHub repository.

Citation

If you use this resource, please cite GenoJEPA:

@article{
  title   = {From nucleotides to semantics: genomic representation learning via joint-embedding predictive architecture},
  author  = {Wang, Chengsen and Qi, Qi and Sun, Haifeng and Zhuang, Zirui and He, Bo and Liu, Siying and Liao, Jianxin and Wang, Jingyu},
  journal = {bioRxiv},
  year    = {2026}
}